E7.6- 10.0,7 a - NASA · RECEIVED NASA STI ACQ. DEC041975 E7.6- 10.0,7 a SEPLAN-CNPq Institute de Pesquisas Espaciais - INPE Sao Jose dos Campos - S.Paulo - Brasil (E76-1CC73) CCILEC-IICN

RECEIVEDNASA STI

ACQ.

DEC041975

E7.6- 10.0,7 a

SEPLAN-CNPq

Institute de Pesquisas Espaciais - INPESao Jose dos Campos - S.Paulo - Brasil

(E76-1CC73) CCILEC-IICN OF .BELEVAKT EESDLTSOE1AINZE WITH THE EFTS-1 SATELLITE IMAGES EYTHE INSTITUTE FCE SEACE SESEAECH (INEE),VCIOME 2 Final Report (Institute dePesquisas Espaciais) 21U p HC $7.75 G3/U3

N76-1356C

OnclasOOC73

https://ntrs.nasa.gov/search.jsp?R=19760006472 2020-04-23T10:24:07+00:00Z

I

I

I

I

I

I

I

I

I

INPE-606-LAFE (Revised)REMOTE SENSING PROJECT (SERE)TYPE III FINAL REPORT

COLLECTION OF RELEVANT RESULTS OBTAINEDWITH THE ERTS-1 SATELLITE IMAGES BY THEINSTITUTE FOR SPACE.RESEARCH - INPE

January 1975PREPARED FOR NASA

VOL.11

CONTAINS trlgliwl jjhotography may be jfordiased from*EROS Data Center

COLOR ILLpHiATIONS 10th and Dakota AvenueSioux Falls* SO 37198

Revised in July 1975 to include correctionssuggested by James C. Broderick - NASA/GSF£LANDSAT Technical Monitor - based on a re-view made by the Scientific Monitor ?

R E C E I V E D

SEP 12 1375SIS 1902.6

\K}|-3* - ix -

TABLE OF CONTENTS

CHAPTER I - GENERAL CONSIDERATIONS

I.I - The Brazilian Space Activities i

1.1.1 -'The SERE Project 2

1.1.2 - The ERTS System 4

CHAPTER II - SEA RESOURCES

1I.I - The Use of ERTS-1 Satellite Imagery in Oceanography 10

11.2 - Rio Grande do Sul Project

Test Site Area 827 - Rio da Prata Estuary and South Coast

of Brazil ... • 17

11.2.1 - Description of the test area ...: 17

11.2.2 - Justification of the area selection 18

11.2.3 -.Objectives of the Project 20

11.2.4 - Coastal Region of Rio Grande do Sul and Rio da Prata

Basin 23

11.2.4.1 - Introduction 23

11.2.4.2 - Interpretation 25

11.2.4.3 - Conclusions 40

11.2.5 - Lagoa dos Patos ;.... 41


11.2.5.2 - Interpretation ,.. 43

11.2.5.3 - Conclusions .. 56

- x -

11.3 - Southeast Coast of Brazil Project •*

Test Site Area 826 - Rio de Janeiro and EspTrito Santo

States and Offshore 62

11.3.1 - Description of the Test Area. 62

11.3.2 - Area Selection Jus tin cation. 63

IL.3.3 - Objectives of the Project 63

11.3.4 - Northeast Region of Sao Paulo State .........' 64

I.I.3.4.1 -General Considerations' 64

II.3.4.2 - Conclusions 70

11.3.5 - Use of ERTS-1 Images in Coastal Studies in the Guanaba-

ra Bay and Adjacent Waters 7]

11.3.5.1 - Introduction i.. 71

11.3.5.2 - Techniques used in the Analysis . 75

11.3.5.3 - Results and Discussion 76


11.4 - Barra do Rio Amazonas Project

Test Site Area 828 - Amazon River Delta and Offshore .... 80

11.4.1 - Introduction , 80

11.4.2 - Interpretation ... 83

11.4.3 - Conclusions -. ^ 88

11.5 - Abrolhos Project ~ .

Test Site Area 808 - Abrolhos Reef 90

11.5.1 - Description of the Test Area go.

11.5.2 - Justification of the Area Selection 90«^f • •

11.5.3 - Summary of the Project " 91

ORIGINAL PAGE ISOE POOR QUALITY;

11.5.4 - Objectives of the Project '. 91

11.5.5 - South Offshore of Bahia State 92


11.5.5.2 - Interpretation ....; . 94


11.6 - Middle Northern Region of BraziliNortheast Offshore Area - Barra do Rio ParnaTba loi

11.6.1 - Introduction ...» 101

11.6.2 - Interpretation .103

11.6.3 - Conclusions 107

11.7 - Bibliography 108

CHAPTER III - MINERAL RESOURCES .

111.1 - Introduction . m

111.2 - Materials and Methods 113

111.3 - Discussion of the Results . 117

111.3.1 - Sao Francisco River Basin 118

111.3.2 - Amazon Region 119c

111.3.3 - Structure Project 121

111.3.4 - Other Areas 122

111.4 - Conclusions 126

111.5 - Acknowledgements .;•;«* 128*• • ••-AAnnex III.l - ERTS-1 Frames for the Different PÔpts 129

Appendix I I I . l - INPE-395-LAFE - Geology of the Areas of the

ORIGINAL PAGE ISOF POOR QUALTTXI -

- xii -

Upper Sac Francisco Basin'and Fiirnas Dam(Erasil). .

Based on Interpretation of ERTS-1 Imagery

Appendix III.2 - Remote Sensing Applications for Geology and

Mineral Resources in the Brazilian Amazon Region -130b

Appandix 111.3.- Geologic Map of the Pre-Cambrian of the Amazoni

Region 130c

CHAPTER IV - SOIL RESOURCES ' . .

IV. 1 - Introduction 132

IV.2 - Preliminary analysis of ERTS-1 imagery with special

reference to Agriculture and Forestry ". 133

IV.2.1 - Introduction ' 133

IV.2.2 - Methodology 133

IV.2.3 - Image E-1047-12274 - ParaTba River Valley Region 133

IV.2.3.1 - Introduction ; 133

IV.2.3.2 - Available Information 135

IV.2.3.3 - Conclusions 135

IV.2.4 - Image E-l054-13070 - Campo Grande Region 133

IV.2.4.1 - Introduction j 138X



IV.2.5V Image E-1105-12532 - Santa Maria Region 141

IV.2.5.1 - Introduction 141



V- xiii -

IV.2.6 - Image E-1048-12282 - Teresina City Region 145

IV.2.6.1 - Introduction i 145



IV.2.7 - Image E-1123-12510 - Presidente Prudente Region 152

IV.2.7/1 - Introduction .v 152


IV.2.7.3 - Conclusions I '... 159

IV.2.8 - Image E-l247-1247-12402 - Piracicaba City Region 160



IV.2.8.3 - Conclusions '. 162

IV.2.9 - Image E-1054-13073 - Dourados Region 165




IV.2.10 - Image E-1048-12.321 - Tres Marias Darn Region ... 169



IV.2.10.3•- Conclusions 171

IV.2.11 - Conclusions ". 173• * c

IV.3 - Mapping of Natural Vegetation Distribution over Central

Eastern Brazil from Data Obtained by ERTS-T 174

IV.3.1 - Introduction v 174"\

IV.3.2 - Studied Area and Methodology 176

IV.3.2.1 -Area '. 176

IV.3.2.2 - Methodology .. -j77

IV.3.2.2.1 - Legend 179

IV.3.2.2.2 - Interpretation Key " -J82

IV.3.3 - Results and Conclusions 139

IV.3.4 - Discussion — ; 190t

IV.3.5 - References 192\I

IV.4 - Estimation of Pasture Projects using ERTS-1 Images 199

IV.4.1 - Introduction 199

IV.4.2 - Methodology 200

IV.4.2.1 - Test Site Selection , 200

IV.4.2.2 - Approach .- 201

IV.4.3 - Preliminary Results 204

IV.4.3.1 - Preliminary Interpretation 204

IV.4.3.1.1 - Channel 7 204

IV.4.3.1.2 - Channel 5 208

IV.4.3.1.3 - Cities -. 210

IV.4.3.1.4 - Roads: 211

IV.4.4 - Other Considerations 211

IV.5 - Identification of Forestal Coverage through the

ERTS-1 Images « 212

IV.5.1 - Introduction . 212

IV.5.2 - Materials and Methods 212

IV.5,2.1 -Materials ., 212

IV.5.2.2 - Methods 213

IV.5.3 - Results '.. 214

' xv '

IV. 6 - Orbital Images Uti l izat ion (ERTS-1 and SKYLAB) for .

Pedological Surveys ............. : ....................... 217•

IV. 6.1' - Introduction .......... . ............. .................. 217

IV. 6. 2 - Test Sites ............ . ....... ; ...................... .. 217

IV. 6. 3 - Materials and Methods ................ „ ................. 220

IV. 7 - Study of the Changes in the Land Use Associated with the

. "Transamazonica" Road Development, using ERTS-1 Images .. 221

IV. 7.1 - Introduction ...... .................. .................. 221

IV. 7. 2 - Methodology ......... ................ ................... 222

IV. 7. 3 - Activities already done .* ........... ................. , 223

IV. 8 - Pastures Evaluation by ERTS Imagery• '

. (Multi spectral Remote Sensing Application in Rangeland

Capability Evaluation for Grazing) .................... 223

IV. 8.1 - Introduction ........................ .., ........ ....... 223

IV. 8'. 2 - Methodology ......................... ------ ......... ..... 225

IV. 8. 3 - Interpretation .......... ........... . ............... ... 226

IV. 8. 4 - Results . .............................................. 231

IV. 8. 5 - Conclusions ........ . ........... ................. ...... 234

IV. 8. 6 - Further Applications ..... ....... ....................... 235

IV. 8. 7 - References ..................... ............... ..c ...... 237

CHAPTER V - GEOGRAPHY

.V.I - Activities of the Geography Group related to the Use of

ERTS Images 239

ORIGINAL PAGE ISPOOH -QUAUTXJ - '

Vol. - xvi -

V.2 - Geomorphological Mapping of the Upper Sao Francisco River. 239*

V.2.1 - Introduction 239

V.2.2 - Materials and Methods 241

V.2.3 - Interpretation Criteria 243

a - Drainage System 243

b- Structural Characteristics 243\

c - Information on Morphoclimatic Systems 243

d - Altimetry 244

V.2.4 - Results 244

A - The Plateaux and the "Serra do Espinbago" 245

1 - Relief with Parallel Crests ; 250

2 - Intermountainous Depressions 254

B -Sedimentary Plateaux 255

C - Tabl el ands 256

D - Intruded Valleys in.the Tablelands 257

V.2.5 - Conclusions . 258

V.3 - Hydrographic Map Using ERTS Images 258


V.3.2 - Drainage: General Organization 259

V.3.3 - Conclusions * 262

V.4 - Demographic Inference Using ERTS Images ... 262


V.4.2 - Methodology 264

¥-*,4.3 - Discussion of the Results 267^ - . . . .

V.5 - Bibliography '.". 280

ORIGINAL PAGE 13 -OF POOR QUALEHS ^ -

" xv11 -

LIST OF FIGURES

CHAPTER I - GENERAL CONSIDERATIONS

Fig. I.I - Matrix. Organization Chart - Remote Sensing Project.. 7

Fig. 1.2 - Map of Brazil with location of ERTS images received

up to December 1974 .. ............................ .. ' g

CHAPTER II - SEA RESOURCES1

I I.I - Shows .the attenuation of light by sea water, as a

function of materials in suspension ............... ... n

II. 2 - Shows the absorption of sea v/ater on the ERTS-1 MSS

channels .......... . ................... ...... .......... 13

II. 3 - Image E-1103-12.415, 3 Nov. 1972, channel MSS-5, scale

- 1:1,000,000 ................................ . ......... 24

II. 4 - Partial reproduction of the nautical chart of DHN (Di-

retoria de Hidrografia e Navegacao - Brazilian Naval

Hydrographic Office) n9 90, scale 1:990,526 (1965), lat. .

031°25'S ... ..................................... ..... 26

II. 5 - Image E-1103-12. 415/5, scale 1:1,000,000; overlay C4,C5

and detail of nautical chart of DHN,n<? 90 ... ......... 28

II. 6 - Image E-1103-12. 415/7, scale 1 :1 ,000,000; overlay C7.c

Note that in this channel the lakes are very well

defined ..... . ........................................ 31

II. 7 - Partial detail of aerial navigation chart USAF-ONEQ-28,

.^ scale 1:1,000,000 and predominant direction of displaced

coastal dunes in the South of Brazil .......... ... ..... . " 33

11.8 - Areas covered by ERTS, NIMBUS IV and APOLLO satellites,

corresponding to the coastal region of Rio Grande do Sul 35

11.9 - THIR of Niabus III, shov/ing ground changes 36

11.10 - Systematic analysis including images of NIM3US IV, THIR

(11.5u) showing approximate limits of Brazilian Current

according to Tseng Yun Chi (1974) 38

11.11 - Interpretation of APOLLO 8 Image - Lagoa dos Patos 39

11.12 - Reduced Photomosaic, channel MSS-7 - Lagoa dos Patos,in

26 June 1973 images 43

11.13 - Isobath - Lagoa dos Patos 46

11.14 - Detail of 45 and 46 sheets (Porto Alegre and Lagoa Miriro)

extrapolated from International Chart of the world to

the millionth (IBGE-1972 edition) - scale: 1:1,000,000.. 47

11.15 - Scheme of coastal lagoons morphologic evolution, associ-

ated to the sandbanks formation, according Zenkovitch

(1967) ..' 49

11.16 - Surface currents distribution in Lagoa dos Patos 50

11.17 - Bathy-hipsometric map of Lagoa dos Patos,/indicating the

vertical profile orientation (DHN.n? 2140) 52

11.18 - Diagram of vertical profile elaborated over the DHN° chartV

n92140 54

11.19 - Schematic outline of the internal circulation cells in La^

r****r.. goa dos Patos Basin 55

11.20 - Image E-1338-12.475 channel MSS-5, 26 June 1973, received .

. . by Brazilian Recording and Reception Station located in

Cuiaba (Mato Grosso), scale 1:1,000,000 57

/. v- xix -'

11.21 - Surface v/aters distribution and circulation in Lagoa dos

Patos, proposed by interpretation.of E-l338-12.475 image,

• channel MSS-5, in 26 June 1973 ...." 58

11.22 - Image E-1338-12.475 channel MSS-6,. 26 June 1973, received

by Station located in Cuiaba (Mato Grosso) scale 1:1,00,0X1 59

11.23 - Surface waters distribution and circulation in Legoa dos\

Patos, -proposed by interpretation of E-1338-12.475 image,

channel MSS-6, in 26 June 1973 60

11.24 - Detail of nautical chart n? 1600, ellaborated by DHN

(Brazilian Naval Hydrographic Office) 65

11.25 - Image from channel MSS-5, obtained by ERTS-1 in 8 Sept.

1973 •. 66

11.26 - Guanabara Bay (Nautical chart n9 1501 frora Brazilian Navy

Hydrographical Office) 72

Sampling location.

11.27 - Sources of pollution in Guanabara .Bay:1. Organic 3. Sewage2. Industrial 4. Docking activities 74

11.28 - Spectral reflectance at Station B 77

11.29 - ERTS-1, MSS-5, Probable Surface Circulation Patterns 79

11.30 - Photographic Reduction of the mosaic of the Amazonas Oirt

fall region made up to MSS - Channel 4 images 84

11.31 - Guide diagram for identification of the mosaic of figure

11.30 85

11.32 - Interpretation of the mosaic included in figure 11.30... 86

11.33 - Percentage distribution of fluvial surface v/aters and

isohalines of the surface in April.according Magglioca.. 87

ORIGINAL PAGE QnpPOOR ^

•xx -

11.34 - Cartographic detail of the sheet SE-24, Rio Doce (36) of

the International chart of the World to the millionth

(IBGE-1972)'about the coastland of the Bahia State 95

11.35 - Partial detail of images E-1224-12.095 in the Multispec-

tral channels (MSS) 4, 5, 6 and 7 96

11.36 - Scheme indicating the order of the montage of the images

of figure 11.35 ....... '. 97

11.37 - Preliminary photointerpretation on the 'images components

of f igure 11.35 montage ....= 98

11.38 - ERTS 1048-12.273 image - channel 4 Northeast Offshore

Area --Barra do Rio ParnaTba ...i 102

11.39 - ERTS-1048-12.273 image - Preliminar interpretation on

the channel HSS-5 105

CHAPTER III - MINERAL RESOURCES

Fig. III.l - First priority area for geological applications of

ERTS-1 data 112

Appendix 11 I.I . . ~ ..-„...

Fig.l - Portion of the electromagnetic spectnsn used in remote

sensing .— 6

Fig.2 - Schematic ERTS-1 configuration .„ 7

Fig.3 - MSS Scanning arrangement 9

Fig.4 - Ground scan pattern for a single MSS detector 9

.Fig.5 - Typical ERTS Daily ground trace (daylight passes only)

ORIGINALOF POOR QUALITY

- xxi -

Fig. 6 - Geologic works in the Upper Sao Francisco Basin Area. 24#

Fig. 7 - Geologic 'works in the Furnas Dam Area 35

Appendix 111.2i '\

Fig. 1 - ERTS-1 band 5 image for the Serra dos Carajas region,

Para State, Brazil 9

Fig. 2 - ERTS-2 band 7 image for the Serra dos Carejas region,

Para State, Brazi 1 » 10

Fig. '3 - Side looking airborne radar mosaic for the Serra dos

Carajas, Para State, Brazil IV

Fig. 4 - Geologic Map of the Serra dos Carajas region obtained

by interpretation of EP.TS-l images and field data .".. 12

Fig. 5 - ERTS-1 band 5 image for the northern part of the Rorai_

ma Territory, Brasil ..13

Fig. 6 - ERTS-1 band 7 image for the northern part of the RorajN

ma Territory, Brazil 14

Fig. 7 - Side looking airborne radar mosaic for ths northern

part of the Roraima Territory, Brazil 15

Fig. 8 - Geologic map of the northern part'of the Roraimac

Territory, obtained by interpretation of ERTS-1, images

and field data 16

Fig. 9 - ERTS-1 band 5 image for the Rio Fresco region, Para .

State, Brazil 17

Fig.10 - EPJS-1 band 7 image for the Rio Fresco region, Para

State, Brazil ••••18

- xxii -

Fig. 11 - Side looking airborne radar mosaic for the Rio Fresco

region, Para State, Brazil • : f 19

Fig. 12 - Geologic map for the Rio Fresco region, obtained by

interpretation of ERTS-1 images and field data 20

CHAPTER l\' - SOIL RESOURCES

Fig. IV.1 - Brazil map showing th'e analyzed frames 134

Fig. IV.2 - Thematic map of the ParaTba River Valley ... 136

Fig. IV.3 - ERTS image E-l047-12274 - channel 5 - ParaTba River

Valley 137

. Fi.g. IV.4 - ERTS image E-1054-13070 - channel 5 - Campo Grande

Region 139

Fig. IV.5 - Map over the channel 7 image on the scale

1:1,000,000 142

Fig. IV.6 - Land use thematic map - Santa Maria Region - origin^

' al scale 1:1,000,000 143

Fig. IV.7 - Interpretation of ERTS images with the support of

Radar Mosaic and Infrared False Color images -149

Fig. IV.8 - ERTS image E-1048-12282 - channel 5 - Teresina

Region •. 150

" Fig. IV.9 - ERTS image E-l123-12510 - channel 7 - Presidente

Prudente Region 153

Fig. IV. 10 - Map done over channel 7 image enlarge-ment - Area 1

Tiete River 154

Fig. IV.11 - Map done over channel 7 image enlargement - Area 3 •

near the. .'-'artinopolis City - Sao Paulo State 156

V o|, V - xxiii -

• '

Fig. IV.12 - Map done over channel 5 image enlargement - Area 2

Feio River Basin Area ' 157

•Fig. IV.I3 - Map done over channel 5 image enlargement - Area 3

Martinopol is Area - SP IDS

Fig. IV.14 - Thematic map of Piracicaba Region .; 163

Fig. IV.15 - Thematic map done over an enlargement of channel 5 164

Fig. IV.16 - ERTS image E-1054-13073 - channel 5 - Dourados

Region , 166

Fig. IV.17 - ERTS image E-1054-13073 - channel 7 - Dourados

Region 167

Fig. IV.18 - ERTS image E-1048-12.321 - channel 5 - Tres Marias

Dam Region ..- 170

Fig. IV.19 - Thematic map of Tres Marias Dam Region 172

Fig. IV.20 - Location .of the studied area 1Q6

.Fig. IV.21 - Average spectral reflectance curve of 240 spectra

from vegetation and 154 spectra from air dry soils 197

Fig. IV.22 - Natural vegetation map 198

Fig. IV.23 - ERTS image E-ll07-12593 - channel 5 202

Fig. IV.24 - ERTS image-E-1377-12584 - channel 5 203

Fig. IV.25 - Land use mapping based on ERTS imageries ......... 205

Fig. IV.26 - Delineation of Drainage Patterns based on ERTS"

imagery 206

Fig. IV.27 - ERTS image 1388-12205 shows the test site location 215

Fig^JV.28 - Thematic map of the Ipatinga Test Site 216

Fig. IV.29 - Test Site locations , 219

- xxiv -

Fig. IV.30 - Leaves reflectance curves ....' 227

Fig. IV.31 - Combinations betv/een plant reflectance and filter

transmissions along the v;avelength 228

Fig. IV.32 - Multispectral photos of the same area in four film/

filter combinations 232

Fig. IV.33 - Rangeland capability ranp •. 235

CHAPTER V - GEOGRAPHY

Fig. V.I - Localization of the Belo Horizonte Chart in the Bra-

zil map 240

Fig. V,2 - ERTS image n9 E-1048-12314 - channel 5 *-..'... 246

Fig. V.3 - ERTS image n9 E-1048-12314 - channel 7 247



Fig. V.6 - Localization of the studied area in the Belo Horizoji

te Chart 251

Fig. V.7 - Area corresponding to the images n9 E-1048-12314 and

n9 E-1389-12255 252

Fig. V.8 - Geomorphological outline of the Upper Sao Francisco

River Basin Region 253

Fig. V.9 - Hydrographic Map of the Upper Courses of the Sao

Francisco and Jequitinhpnha Rivers 260

Fig. V.10 - ERTS image n9 E-l372-12333 - channel 5 . 265

Fig. V.ll - ERTS image n9 E-l048-12330 - channel 5 266

- XXV -

Fig. V.12 - Brazil map with the localization of the studied0

cities 268

Fig. V.13 - Dispersion Diagram 276

F'ig. V.14 -.The vertical bar that represents the probability of

90% of the population interval for a gfven class

(•area) ..." 27S

Vol. • Xxv1 -

• ' LIST OF TABLES ...


Appendix 111.1

1 - Comparison among the four bands ........................... 19

2 - Correlation of the remote sensing units to the geologic

mapping units in the Upper SacFranciEco Basin ... .......... 30

3 - Correlation of the remote sensing units to the geologic

mapping units in the region of Furnas Dam ................. 37

CHAPTER IV - SOIL RESOURCES-

Table IY.1 - Legend .............. ---- . ..................... .. 194

Table. IV. 2 - The training areas for establishment of the inter-

pretation key .......................... .. ........ 195

Table IV. 3 - Flight Parameters ...... ............... ........... 229

Table IV. 4 - Summary of Rageland Capability for Grazing ....... 236' •

CHAPTER V - GEOGRAPHY . .

Table V.I - 278

.-•4**. .

>|, V - xxvii -

LIST OF PLATES


Appendix III.I

1 - ERTS-MSS Imagery n9.1048-12321-5 with geologic interpretation

overlay 38

2 - ERTS-MSS Imagery n9 1048-12314-5 with geologic interpretation

overl ay 39


overl ay .." 40


overl ay 41


overl ay . 42


overlay 43

- 131 -

CHAPTER IV

SOIL RESOURCES

Principal Investigator;

Mendonga, Fernando de

GSFC ID-F0398 MMC 326-01

- 132 -

CHAPTER IV- - - _ *

SOIL RESOURCES

IV.1 - INTRODUCTION

The Soil Resources Group has used the MSS/ERTS-1 images

since October 1972. It has tried to conduct researches in accordance to

the Brazilian proposal to NASA for utilizing ERTS data but also keeping

in mind the economical importance to define the priority of such studies.

First of all, the ERTS images have been analyzed for

general purpose of surveying agricultural features. After that, the ERTS

images were used in more specific researches such as natural vegetation

mapping, forest inventory, deforestation control.

Other researches using ERTS and SKYLAB images are being

carried out including soil maps, studies of crop forecast data and

rangeland management.

The following pages present in more details the work

the Soil Resources Group is dealing with.

- 133 -

IV.2 - PRELIMINARY ANALYSIS OF ERTS-1 IMAGES WITH SPECIAL

REFERENCE TO AGRICULTURE AND FORESTRY

IV.2.1 - Introduction

The purposes of this work are to define the ERTS-1 images

potentiality for agricultural surveys and to provide basic information

for further research . Owing to the short period of study and the absence of

ground truth data, the results reported here are therefore of a preliminary

nature. We call attention that some vegetation types are described in

item IV.3.2.2.1.

IV.2.2 - Methodology

The preliminary analysis presented here was basically done

using black and white photographic copies, on the scale 1:1,000,000 of the

four ERTS-1 channels. However, some particular cases appear where magnified

images were used as well, as other remote sensing information sources, in

addition to the ERTS-1 images.

Figure IV.1 shows the Brazil map where the ERTS images were

numbered to put in evidence the frames under study.X

IV.2.3 - Image E-1047-12274 - ParaTba River Valley Region

IV.2.3.1 - Introduction

In this analysis use was made of the above mentioned image

- 134 -

2. CAMPO GRANDE

7. DOURADOS

5. PRESIDENTS PRUDENTE--

3. SANTA MARIA

4. TERESINA

— 8. TRES MARIAS

— 1. VALE DO

— 6. PIRACICABA

Fig. IV.1 - Brazil map showing the analyzed frames.

- 135 -

with image center coordinates 23°06'S and 45°17'W, dated September 8,1972.

This image embraces the ParaTba River Valley, "Serra do Mar" and part of

the Paulista coastland. This region varies greatly in topography, climate

and vegetation. Cattle breeding is a major activity in the area.

IV.2.3.2 - Available Information

1 - Based on the analysis a thematic map was done .(Fig.

IV.2) which could be compared with the channel 5

image (Fig. IV.3). The boundary establishment of the

mapped units was done over the channel 5, to present a

better contrast, although the identification was done

also using the other channels .

2 - the landforms aspects are better evidenced through

channels 6 and 7.

3 - Vegetation information are more easily obtained using

, the channel 5 image where the variations from dark to

light gray tones indicate a graduation from dense

vegetation (forests) to sparse vegetation, corresponding

to areas with herbaceous species and grassland.

•v

IV.2.3.3 - Conclusions

1 - The information acquired from the orbital ERTS-1 images

are of. great importance to plain agriculture activities

because it permits a previous knowledge of the area,

when speaking in a global way.

- 136 -

DENSE VEGETATION (FOREST)

DAMSRIVERS

FOREST ON THE "SERRA DO MAR"(MAR RANGE) SCARPSHIGHWAY AND ROADS

/'V;N PARAlBA RIVER VALLEY\ v V •/-i-s

Fig. IV.2 - Thematic map of the ParaTba River Valley.

- 137 -

II

-~ - -01 U005-00I IS02a-0038SEP72 C S23-06/UW5-I7 N S23-05xWM5-10 HSS 5 ' R SUN Q-«3 RZ056 188-0652-fl-1-M-D-2L MP.SR ERTS E-10^7-1227^-5 01

Fig. IV.3 - ERTS image E-1047-12.274 - Channel 5 - ParaTba RiverValley Region.

- 138 -

2 - Based on a summary analysis of another image set for

this same area, taken at different ERTS-1 coverage

periods, it is possible to accompany the evolution of

agricultural activities (crops evolution).

The ERTS-1 repetitivity permits a better forest

monitoring due to the official government policy

stimulating and controlling reforestation.

3 - It is important to point out that the thematic map

presented here was obtained using only the MSS/ERTS-1

images, without additional costs for field works; besides

it was done in a short time.

IV.2.4 - Image E-1054-13070 - Campo Grande Region


The area selected is located in the center-west part of

Mato Grosso State, with image center coordinates 20°19'S and 54°34'W,

taken in September 15, 1972. It shows the Campo Grande City in the

image center.x


1 - The cities of Campo Grande, Bom Fim and Sidrolandia

were easily located through channel 5 (Fig. IV.4). In

this same channel the highway networks specially the

- 139 -

'.*'~f-rtr ?.f

•

. . , . .' N ' >'

./-*• \\rC -C— S

SIDROLANDUV-,.' ^

Fig. IV.4 - ERTS image E-1054-13.070 - Channel 5 -'campo Grande Region.WET AREA .££WET MIXED WITH DRY AREA

$ |D49"sl6>fWrCHANNELS 5 AND.7

i.'aiaf -UNITS

- 140 -

Cuiaba - Campo Grande and Campo Grande - Sao Paulo

highways can be identified.

2 - Channel 5 shows two drainage systems oriented in

different directions, one Southwesterly directed and

the other to the West.

3 - Several vegetal units and land use are shown in the

transparency over the figure IV.4 as:

unit 1: Area occupied with natural vegetation, typical

of Central Brazil, named "cerrado". It appears

in light gray tones to the East and Southeast

of the area indicating soils which are more

eroded. The left side of the image appears in

darker gray tones, evidencing denser vegetation,

with well integrated drainage pattern, with an

erosion degree from light to moderate.

unit 2: Area with "cerrado" vegetation including

pasture, situated at a higher level and which

at a first sight appears as areas without

agricultural uses.

unit 3; It presents mottled gray tones indicating strong

agricultural activities related to crops and

pasture. This is more evidenced in the vicinities

of Campo Grande city expanding toward Cuiaba city.

4 - Channels 6 and 7 show areas in dark gray tones and with

well defined geometric forms indicating irrigated rice

crops.

- 141 -


1 - The areas with agricultural activities appear

surrounding the urban centers.

2 - Some areas are suffering a strong erosion action

needing some soil conservation measures.

3 - There exist two drainage systems. One in direction to

the Paraguai River and the other, at the opposite side,

in direction to the Parana River.

4 - The area does not present a good transportation routes'

system. Just some railways and few paved roads.

5 - This region possesses great variety of soil types

indicated through the natural vegetation and land use

differences. In the areas to the West the soils trend

to be better.

6 - There are great quantity of rivers in the area which

permits one to say that a well-planned agriculturali

activity would not suffer from water shortage.

7 - The relief of this area is flat, presenting itself at

the Northern and Northeastern regions a little more

irregular.

IV.2.5 - Image E-1105-12532 - Santa Haria Region


The above referred image (center coordinates 30°17S-53 03W),

- 142 -

.-•>•-.. :•*'• •.. .\'' *

7 .N X

C A C A P A V A

SUL

DO

0

>»Lakes, Sluices and flood areas. ' .

Fig. IV.5 - Map done over the channel 7 image (scale 1:1,000,000)

- 143 -

Fig.IV.6 -

- SedimentaryBasin withpasture andannual crops

Land use thematic map - Santa Maria Regionoriginal scale 1:1,000,000.

- Cloud coverage

- 144 -

obtained at 5 November 1973, was analyzed considering the soil plowing

and planting period and the begining of some annual crops growing. The

image shows the cities of Santa Maria, Cachoeira do Sul, Cagapava do

Sul, etc.


Information were obtained using, besides the four channels

on the scale 1:1,000,000, an enlargement of channel 5 on the scale 1:500,000

and one color composite of channels 4,5 and 7.

1 - Utilizing the channels 6 and 7 it was possible to map

lakes and sluice distribution along the right margin of

the JacuT River (Fig. IV.5).

2 - A thematic map (Fig. IV.6) was obtained through a combined

analysis of the four MSS channels on the scale 1:1,000,000.


1 - It is possible to count the number of lakes and sluices

- as well as their superficial area using a channel 7

photographic enlargement.

2 - The great concentration of water bodies along the JacuT

River determines the great occurrence of rice crops.

3 - It was easy to separate the following categories:

annual crops, pasture and forests.

- 145 -

4 - The outline of woodland, remaining in the area .could

be delimited .

5 - The analysis of image sets, taken at different dates.

through the different growing crop cycles, with regional

specialists support and also with aircraft sampling, at

intermediate scales over test sites areas previously

selected on the ERTS images, would permit more detailed

results about the present and potential land use in

this region.

IV.2.6 - Image E-l048-12282 - Teresina City Region

IV.2.6.1 - Introduction*

The image under analysis has the central coordinates: 05°49'S

and 42°28'W, taken at September 9,1972. It includes the city of Teresina

and parts of the ParnaTba and Poti Rivers. To the West of the ParnaTba

River appears the Maranhao State and to the East part of this river the

PiauT State.

The vegetation in this area is predominantly natural

varying from small and sparse "cerrado" vegetation, used as pasture, to

"cerradao" (vegetation of forestal size) with some agricultural activities.

It also presents in its lowlands and river beds the "babacu" vegetation,

. - 146 -

one of the main sources of vegetal oil extration of that region.»


Information were obtained not only using ERTS images

(1:250,000) but also from infrared photographies enlargements to the

scale 1:130,000, SLAR (RADAR) mosaic on the scale 1:250,000 and previous

field observations from part of this region.

The analysis of these images could be summarized through a

thematic map as shown in figure IV.7 which was drawn over the channel 5 of

the ERTS image (Fig. IV.8).

Several types of features were plotted on the ERTS image

and in the case of some of them the RADAR image correlated quite well,

while other features were not perceptible in the radar such as those

microcosms with boundaries in tracing.

Basically the following were the types of vegetation and

soil use established through the images:

Area 1 - Area 1 is characterized by a rather dense savanna vegetation,•x.

of the "cerradao" type, with a large number of cultivated areas.

It presents good agricultural activity in spite of its rather

accentuated relief. A patch, in the shape of horseshoe near

the parallel of 6°S, crossed by the high-voltage power line

- 147 -

of the "Boa Esperanca" Hydroelectric Company (COHEBE), ha's been

Identified. Both in the RADAR and in the infrared images cultivated

areas are found near this horseshoe. In the ERTS image this area is

unfavourably located, too near the left edge, and this resulted

in blurred details.

Area 2 - This area is characterized by "cerradao" vegetation and small

agricultural activity, which is due to relief (shown by vertical

arrows) as a limiting factor. In this microcosm we find yet

another area well characterized in the ERTS and in the RADAR

images (area 2.1). Patches F and G, which are not well delimited

in the RADAR image, but are well characterized in the ERTS images,

belong to the type 4 vegetation. The vertical arrows indicate the

most hilly part of the area.

Area 3 - Areas characterized by number 3 classified as "cerrado" vegetation,

that is, a less dense savanna type identified by a fine texture

in RADAR and a medium gray tone in the ERTS images. The index of

each area indicates the variation of their respective reliefs as

well as the parts where agricultural activity can be inferred.

Area 3.1 - This area shows parts covered by "cerrado" with flat relief as

well as great agricultural activity in other parts. It comprises

the environs of FEITORIA, AGUA BRANCA and SAO PEDRO DO PIAUl.

Cultivated areas can be perceived both in the RADAR as in the*

infrared images. The texture of this area in the RADAR is well

- 148 -

characterized (fine texture) and the relief is smooth.

Area 3.2 - The 3.2 areas differ from the foregoing only by the higher relief.

Area 3.3 - Areas in 3.3 show the same characteristics as the previous two,

differeing only in the degree of certainty as to the existing

agricultural areas. The area as a whole is of agricultural

potential.

Area 4 - Microcosm number 4 comprises a large part of the wuadrilateral

under study and is characterized by vegetation so sparse that

the reflectance of the soil has influence on the light gray tone,

characteristic of vegetation in the ERTS image under study. In the

RADAR image individual areas are not well characterized in terms

of relief and texture but some boundaries have been plotted which

allow comparisons.

The whole number 4 area can be considered, in a general way, as

sparse savanna unfit for agriculture.

i

Area 5 - The area delineated as n° 5 is similar to area n9 1, but shows

well the agricultural areas near the poti River. Sub-area 5.1 is

well evidenced, in the ERTS image, as a darker gray spot.

Predominantng vegetation in the proximity is of "cerradao" type of

almost like it. Area indicated by arrow, near this region,

probably belongs to group 4 and is perfectly characterized in

all the images.

The spots marked with letters /\ to IH have sparse vegetation, with

varied relief and no agricultural activity.

erts • 1:1000000L E G E N D

1) "CERRADAO" WITH CULTIVATED AREAS

2) "CERRADAO WITHOUT CULTIVATED AREAS

3) "CERRADO" 1 - FLAT LAND AND CULTIVATED AREAS

2 - UNDULATING TO ROLLING LANDS AREAS

3 - WITH POSSIBILITIES OF AGRICULTURAL ACTIVITIES

4) "SPARSE CERRADO" WITHOUT AGRICULTURAL ACTIVITY CONDITIONS

5) "CERRADAO" WITH SEVERAL CULTIVATED AREAS .- '•'6} HUMID AREAS

A THROUGH H - SPARSE VEGETATION, DESCRIBED IN AREA 5.

Fig. IV.7 - Interpretation of ERTS images with the support of Radar Mosaicand Infrared False Color images.

I - 150 -

+

-~^ '* ; %. ^ , v.' - - 'j.< V • • - • » » • . '*> - * ^ vl ' • ?•** *. »- . .. - ,- • ...I c '••»>»%w.v:.v^> ,;ê-i-vi- ,>•'*•;•;.'** -,. ;.. ^^"^'. ,c.Lr--:v;-'.', >,^ ^,. •:?.>..-\v 1- • " • "*• "•-' * • • > / * . .,."'--'•'> .v . - ' V 5 - - v: -*--1'* - •>'-'- '^r:--- r-"-•'-'- '•"• ' '' --"*-~-s

1 ^ - ^ O^n*^X'-;.. ^ ::U->?- ^.v.V^'-V^Vr., V»V.,.? ^-' '.^ . '-• £. ^.;^v%^x^^^v:.^^v-;r/ .;-û-^^^'•**?-•'•:?'-.-.-.-^* '*«•»«, . '.,;;«.-^:«:-/ u,"- - v*-.^,,^^.^,..^.,....,,_, s _,.. «*_,,_ ..;.. •;•- -^ _.. •-. ., _^- .:- •*••• '&•'• .-i'-^Tf J— '. ' '.- . -" ' •• "• ..': t.~ Jf-^ - :•:* . - * . - • .

I

I

Ia

I

I

I

I

i* o ~«^.'•*•'' '!• - N*-- •' '. -I - • ' • " ? .-• « • i •' f- »•• • - . - . • - ---», -^ - • .. • *••"**-<• ' --~r^-T- •-•".'•'*'.^-f< i^'-r- •-,-" ' / \~ " ' - • . - ' • ' • -> • - ' " ' - » , - - ^

» •*•;•'»•«,-3^--:.v;'A- * •-~ii-'i%-';-t: -.VÂ" " ^ -'• ' ' ?.'* ---'• •"• •••' - . - - . - --if'-V^' • , -v ' ;-^-^-v-5>>-'f'i->-r > : ^ ' -T^ T.-^ ;>; ;V'V-- ^ * •>. '^'.•:•-;-"•'• -~ >• ^>^-*

U043-00I ' U01J2-30I U842-00I09SEP7Z C S05-49^JB42-28 N S05-50/WW2-23 tISS 5 R SUN EL53 H2072 188-0666-fl-l-ft-D-2L NRSfl ERTS E-l»18-l2282-5 02

Fig. .IV,8 - ERTS image E-1048-12.282 - Teresina Region. Channel 5. "

.-L

- 151 -i

They probably indicate soils of low fertility and appear all over

the image quadrilateral.

Area 6 - These are humid areas which show up well in channel 7. They have

fallen outside of the RADAR mosaic which does not fully correspond

with the area covered by the ERTS image.


1 - Altough a thematic map is presented here only with the

great mapped units, it is possible to obtain a more

detailed level of study on a 1:250,000 scale map using

a channel 5 photographic enlargement.

2 - The unit 1 area, in spite of being intensely cultivated,

was not clearly evidenced im the MSS images due to the

reduced size of the planted areas located closely together

and from the same stage of growth, reducing in this way

the contrast among them. In area 2, channels 4 and 5

show the existence of agricultural areas.

3 - Through this preliminar interpretation it was not possible

to separate the "babacu" vegetation from the "cerradao"

neither at the false color infrared images, due to the

. inferior quality of the iirages.

4 - ERTS channel 5 images to the scales 1:1,000,000,

1:500,000 and 1:250,000 allow a good characterization of

large vegetation types of the area under study. A good

subsidy in relation to relief and texture was given by

the RADAR 1:250,000 images. The joint use of the two

- 152 -

types of images makes for a much better interpretation•

than the use of either one of them in isolation.

5 - The additional information obtained from the 1:130,000

infrared images was of great importance in this stage,

1n which interpretation techniques is not yet fully

developed. The infrared images gave details not

perceptible in other two types of images studied, but

unfortunately with bad quality in terms of cloud

coverage and color enhancement.

6 - Water bodies are equally well characterized both in

ERTS channel 7 as in RADAR images. Relief shows up

better in the RADAR image but geometrical accuracy is

better in the ERTS images.

IV.2.7 - Image E-ll23-12510 - Presidente Prudente Region


The analysis referred to the above mentioned image with

center coordinates 21°36'S and 50°42'W taken at Nov. 23, 1972, corresponds

to the region of Presidente Prudente, MarTlia and Andradina cities located

.at the Sao Paulo State.


1 - In extense areas along the Tiete River it is possible

to observe a defficient drainage evidenced by the dark

- 153 -

U35I-G8I u^Da-aai

X'\l££.v77•r-?V^;.-'.'-:f^"..- '-f^|

-j. ,r-/>: ^•&&Zf*$fw%ft<3&%ff*?ir:^n^^':^^-^m^mr. % I > * • • • - / ' • • •- \ .««- . JL -^ _ .. . « IM .k. * • "i. ' « .. ._ —'^rmffr

''•'3^$^$fy:$&?g£$f- ':• - " ' • • • • '•'«^V;:''v V.'V- -:' -" v\- '*;-;-'•:• "•:'- ;-.f ;*-*r* •. .• -,-."A'' 'r ., ^^jir^WV''1" ';"--i-^r' •?-i-rrs^^V^^^'^St'^v ^^ ^^-i'tl-^fcSiî^.:.^:^^^^>«^•OL^^£tî/^•fe>•Ût•^sM^v,v:A:.^>->^

/":" ;:

<t' J* •L. * -""» * '" **• ' " f "• "<"J

^t .».. :'•• '_T^r••••>-^''-

"'-- - : i.V>.- ;x>vv?l;£.;.- ->.', 'v,' ~'-i'.-'•&•:£ ••\.^f'\ -" Jr r ." iw^l

..••-': ^.- -r"'xx^-:-^^3r^^ ^. .i^../.^...^--^.::

'•' ->^^?^û2LS-lLi£i; T^^fv •:: ^^s^w'A!,^.',,.... ,,.,. .f^'P^jvEA :IS;^ffll&îS^^^^"'-: i '-"•-. f^ -^*>'r^;'-^l:'i^-<^V:>'ii:A^.S:>^ r:'-)/-;';:"*'^V'?-':'>'--^!ii"";';"'.•

V' -*.^--.^--^--^:^fe--:;-.^ -.^îX'VWR,.-, -: Cp*^X|J^^^Ê?Tniri^^

._ , '-r.--.~v • • • -

~ ' * • * '

AREA'•'•,'"'''• v'i -' "*J^ •

rx^1 -".•"'•"-;;'•. i-i^'^: • ^ f\-\ -. ;

••> \^'.\.-i^.p-f?.-.»_.-."-;, .•.->X>:\> • ;,•<

IU85I- K35S-38I IS922-38~

iiric I _ r*r» i

23HOV72 C S2l-36/UKe-« N S2I-39^050-36 flSS 7 R SUN EL57 R2094 189- I7I2- 'R-1-N-D-1L >ViSR ERTS E-l IZ3-I25IB-7 «3

Fig. IV.9.- ERTS image E-1123-12.510 - Presidente Prudente Regionchannel 7.

- 154 -

liiwi? 'W&L-.IlifiiWiA. t/^v• / * // * > r < * / • * / / • L*, * / ' * * /^-^~ / / / / R / i i r\ \i i f ** ?*• tJi i /< ;• TT>I ~^/-//////v^/;//// m W^~ ^HBr^r iliiLlPi Vs-

y of

,'!!!!!>>''i!' LHi! if

ii.^±h

* .*'////* .v*i

• * *..» V -

> / / ,' J V

«CUJa:

UJa

«cUJce

«cC£O

oo

.•V/

to

aI

UJ

3UJa:

OO

pi// /.']/ / I l l

i/M

- 155 -

gray tone in channel 7. In this same area a great

number of lagoons is also observed.

2 - Areas of bare soil, of a good drainage, evidenced by

light tones in both 5 and 7 channels.

3 - The transparency done using channel 7 (figure IV.9)

shows the main drainage system orienting itself

parallelly toward the Parana River, to the West of this

image and not shown in the above mentioned figure, as well

as innumerable small lakes, dams and areas influenced by

flood along the Tiete River.

4 - Figure IV.10 shows a map done over the same photographic

enlargement on the scale 1:130,000, of the marked area,

at the upper right part of the channel 7 transparency

(area 1). This area under the influence of the Tiete

River shows lagoons, part of this river as well as areas

close to its river bed with various degrees of humidity.

5 - Figure IV.11 shows the map done over the same channel 7

transparency, in its lower part (area 3), situated near

the Presidente Prudente city. It is possible to note

the presence of two dams and areas of bare soil, with

different moisture degrees, indicating differences in

the drainage capacity.

6 -r Figure IV.12 shows the thematic map done over the channel

5 photographic enlargement on the scale 1:130,000, at

the central part of the image, near the Feio River area

(area 2). In this region the existence of forests areas

and areas prepared for plating, etc. was verified.

- 156 -

:-:*•- i DAMS WATERS

AREAS WITH HIGHMOISTURE CONTENT

AREAS WITHC LOWMOISTURE CONTENT

AREAS WITH MIDDLEMOISTURE CONTENT

Fig. IV.11 - Map done over channel 7 image enlargement - Area 3 near Martinopolis

Sao Paulo State.

Ati/virr.^ PT;\feW 4£/.ft/- fv^V '—<

• \ \ . \ \ \ \ \ \\\ \ \ \ . \ \ \

\ \ \ \ \ \ \ \ -:N» • \ • « \ . \-.* • \ * » *

\ x' v —V V-. \^^—Vv

\"\ \V\ \ \ \ \ \ \V

O-

£% -158- h. rvVx Jo-V-'\ :v \ *

t:~ o-O- o-

CnO O- ->1O-.

c--.«ijf C-

.4, . jV:^/^,

«(r\^ 1 I.f /? c3^ \v—J ^ -'\y ^ (Q>-A li •

• •*— . rn T^. ;• ' r-. • , . .^ v , %. o.:^«^/-/V,&^/•^ ^\V>~ ^-:. ~ r -•-"' ^-. : .• i- ». !•>

fO

2!<cto

oo.

10

iCO

rocc10

^=oi_<uoO)co

Q.to

LU <CCO I—C£. LU«s oQ- LUC0>

•X=^

X LJ£2 CO Zs o: o

<c »-ico a. i—I— co «cco v-LO 3: LUc: i— oo •-< LU

o-

• H '

•—i CD LiJO LU C3CO > <ccLU Z LUan~>-ff.fi OCO 3 CJ

- 159 -

7 - Figure IV.I3 shows a map done over the same photographic

enlargement of channel 5 in the right part of Presidents

Prudente region, at the lower part of the image

(area 3).

IV.2.7.3 - Conclusions1

1 - The area under study does not present great limitations

to the agricultural exploitation practices. This fact

was evidenced through land use, which in its great part

shows some indication of agricultural activities.

However, some areas along the Tiete River do not show

cultivated lands which may be attributed to the

defficient drainage.

2 - It is possible to separate soils according to their

potential use by the different infiltration rate which

they present.

3 - In this area dark spots could be detected and mapped

principally in regions where the agricultural activities

are poor as the ones along the Feio River. Otherwise at

the Presidente Prudente area there exists strong

agricultural practices areas and small areas with

remainder forest.

- 160 -

IV.2.8 - Image E-1247-12042 - Piracicaba City Region

i


This image under analysis is taken at March 27, 1973 with

center coordinates 22°56'S and 48°11'W and includes the cities of Limeira,

Rio Claro, Piracicaba, Botucatu, Avare, Jau, etc. at the Sao Paulo State.

This image was chosen, although presenting two great cloud coverage spots,

because of intense agricultural activities zones and reforesting areas.

IV.2.8.2 -Available Information

Information were obtained using besides the four channels

on the 1:1,000,000 scale, the channels 5 and 7 photographic enlargements

on the scale 1:500,000.

1 - Figure IV.14 synthetically contains the information

related to vegetal coverage obtained through the MSS

multispectral analysis.

2 - In spite of the cloud coverage occurrence disguising

the dams area, their outlines are easily delineated

in the channel 7 image.

3 - The channel 7 enlargement very much facilitates not

only the dams tracing but also rivers outline. The

drainage system is little evidenced what permits to

infer that soils are of reasonable permeability and

- 161 -

porosity and the areas are formed by flat and undulating

relief. Parallelly small lakes and irrigation deposits

for small cultivated areas were mapped.

4 - The road system, even at channel 5, was not well

characterized. Only at a section of the Presidente Pru-

dente highway, between Avare and Campinas is possible

to be seen. To the West of the image the road system

appears with a good integration degree.

5 - During the above mentioned thematic map preparation it

was possible to isolate a category whose texture and

tonality in the image do not approximate that of the

natural or artificial forests, neither of the "cerrado"

vegetation category or pasture. This category was

defined as sugar cane because of its proximity to the

Piracicaba city known as a sugar cane production region.

In this region large cane fields, with other crop

• species appear inserted. They were not mapped due to

the mapping scale.

6 - Figure IV.15 shows a thematic map done over a channel 5

.photographic enlargement on the scale 1:500,000. This

enlargement permitted a better definition of the outlines

of mapped vegetal coverture units and the separation of

new reforested areas.

- 162 -


1 - The analyzed area possesses intense agricultural

activities, flat and undulating relief. The reforestation

and natural forests are concentrated to the West and

Northwest zones of the image.i

2 - For the mapped units characterization in this area

the channels 5 and 7 are the ones which provide more

information.

3 - Through texture type, tonality and man made aspects (as

for instance roads among the plots of land) presented in

the image by both categories, it was possible to separate

areas with natural forests from the one with artificial

forests.

4 - It was possible to identify the following kind of vegetal

coverture (Figs. IV.14 and IV.15): natural forests,

' artificial forests, prepared soils for cultivation,

"cerrado" vegetation areas, pasture and zones with

intense agricultural activities. The last three cathegories

could not be delimited.>

5 - In the natural forests category it was not possible to

separate areas with pinus and eucalyptus, because of the

absence of field information.

»sA'.I- ' '' '•-*• '-*-'&XW/>:> A

'"' r/n-./."^V ^ ..Jf'l^ VA P-< ^* • *"*'", >• .'.'•-•%•'', -"^-••/-/ jî '• •••Cl itj

VT/^^SWÎY ^' 4,f" «?% ' j p^^^^^K*/uB^£,^3 j

^ ^V'/^'^<.;>^v * tffX ••! r- c;-.>--,-.-.^*^?^ S^-v^N^ --IJ^L^ , -u;- r - itfxfei? g

^ ' "^VA*2^ P' f1^ «y;*^7 r-, £*f*-v*« / V? b!

y^V. V V V ' » ' \ A * -i

V.v.v:^/' ^^

. ^v'-lH'fi*^j-îrvi- ^y^--,.^:, t:*>t^^-^vi A v 'r'vy" ^""^h,:Ki,ts \ r ¥^7^h tz, |Uf W-"-'!- -, Q \ \

^^ «fe-»S^ %, XJl X'"\ro* y/:ijv:*Jv> tii? rz1^*^-*^ , 1

A EOTUCATU

O»

x\

e TATOI

NU V E - N S

' ^ NATURAL VEGETATION

REFORESTED AREAS

AREAS PREPAREDFOR CULTIVATION(BARE SOIL)

GALLERY NATURALVEGETATION

oSUGAR CANE

ROADS

LIMITS OF THECLOUD COVEREDAREAS

DAMS,RIVERSAND LAKES

A - ZONES WITH INTENSEAGRICULTURAL ACTIVITIES

C - "CERRADOS"

P - PASTURES

Fig. IV.14 - Thematic map of the Piracicaba Region.

- 164 -

2230

,2= ,^ ^%V:J^<%^ *f A *VJB^^ t ~#

* %fe^ ':$> ,4 • •;• (^ \ 4 •

kan 1%??; ^ i".v NJ ',/j '''* ^» • • ,-rr\ ' /ji VH rj' >^

feii>> s«%^ <<3 ^*^ -., ^ «•* .c'Cv^S^ .^» C'-1V^*- ^ -I&/X „ 4^ ^ ^ ^^ • c

r^Vo J& /^ P&»

^ <S^< ,rL ^"^_rJ?^V". x\\^ •;::;? «. m >«.. oi> enrtS- ^^^S^KV,,"' ^1

1 ^>

sfL> !COVUCAYU

/

48«30

^7-.)- NATURAITÊGETATION• vr ' \

• )- REFORESTED AREASW 9 *'J

f^\^ DAMS,RIVERS, LAKES\j*fft RECENTLY REFORESTED AREAS

AREAS PREPARED FOR CULTIVATION(BARE SOIL)

A - ZONES WITH INTENSEAGRICULTURAL ACTIVITIES

C - "CERRADOS"P - PASTURES

o-Fig.:IV.15 - Thematic map done over an enlargement of channel 5.

- 165 -

IV.2.9 - Image E-1054-130.73 - Dourados Region


The above mentioned image, with center coordinates 21°45'S

and 54°56'W, dated September 15, 1972, comprises the South region of the

Mato Grosso State. This region was chosen because of the existence of a

soil map permitting a comparison with the image information.


1 - The drainage system is easily visible in channel 5 as

shown in the transparency over figure IV.16. Generally

the drainage streams appear under gallery forests which

produce dark tones, easily delineated in the image.

2 - There appeared some remainder natural forested areas

evidenced by the dark gray tone in channel 5,principally

to the West and Southeast of the image. The forest South-

westerly directed are regrowing forests (brushwood)c

evidenced by light gray tones.

3 - The agricultural settlement of Dourados appears at the

Southeast of the image.

4 - The comparison between channel 5 with a soil schematical

map showed a good relationship principally in the occurrence

of areas of quartzose sands to the Northeast'of the area and

- 166 -

5322-281I5SEP72 C S2I-<!5/U05a-56 N S2I -•s5/U85a-50 HSS 5 R SUN £Lfl6 RZ360 189-0750-fl-I-N-D-2L r.'FJSfl ERT5 £-135^-13073-5 01

Fig. IV.16 - ERTS image E-1054-13.073 - Channel 5 - Dourados Region.

2-301 IU055-38 W855-00I U054-30II5SEP72 C S2l-«5/U85«-56 N S2I-45/U054-50 MSS 7 R SUN EL46 RZ060 189-0750-H-l-tt-iD-IL NHSR ERTS E-105^-13073-7 0J_

Fig. IV.17 - ERTS image E-1054-13.073 - channel 7 - Dourados Region,

- 168 -

the association of dark red latossol with quartzose sands

in the agricultural settlement area. Although the right

upper corner of the image is not mapped yet, it is

possible to infer the existence of several areas with

quartzose sands.

5 - The image in channel 7, figure IV.17, shows with a sharp

contrast the purple latossol spot (darker gray tones)

and also the great spot of dark red latossol, in lighter

gray tones.

6 - There exists an area in the soil map down in the Dourados

River, near the Agricultural settlement, classified as

purple latossol, but through the image analysis it could

be classified as dark red latossol. However it would be

important to have a field checking.


1 - It is possible to map areas occupied by forests on the

scale 1:1,000,000, as well as areas under cultivation or

prepared for planting, besides the drainage system.

2 - Basing on the already done interpretation over this image,

it is possible to conclude that the utilization of the

several ERTS-1 channels, principally 5 and 7 channels,

for soil mapping purposes, mainly at a schematical level

or even at a reconnaissance level, would be of great

importance. Besides allowing a multispectral vision of

- 169 -*

t

the area, the ERTS images allows the observation of

different soils moisture contents, vegetal coverage,

and a synoptical view of the area.

IV.2.10 - Image E-1048-12.321 - Tris Marias Dam Region


The image under analysis has as central coordinates 18°48'S

and 45°36'W, dated September 9, 1972, comprising the Tris Marias Dam Region

at the Minas Gerais State.

IV.2.10.2 -Available Information

1 - The exact evaluation (high contrast) of the Tres Marias

Dam surface is possible through direct measurements over

channels 5,6 and 7 images, as shown in figure IV.18 -

channel 5. Such information increases its importance

when we consider that ERTS-1 data repetitiveness coulde

--;- be used for accompanying the water volume variations•v

stored during the year. In fact, the comparison done•

between the ERTS-1 images (taken at the end of the dry

season) and the USAF map, based on photographies obtained

in March 1964 (end of the rainy season), clearly shows

differences in stored water volume inferred by the

variation in the dam outline. The ERTS-1 repetitivity

IIII

- 170 -

W0<»5-30I S018-081

*Si!&^&*&&®%SF*

' ^J .J^^ '-^^^#^-^?' - • ' „ •-- - . ' .' '- - .- ' -V • • > . . - -•.'• . '':> ?*•">•- • < . - - : * '-A ~ ' V "

••"'••-*» - "'-' V/f' -C-• "•"'U -?-V ' • /V J-}-1^ /*? •"'"" """ '"•"• '- • '"' i ,v* -'«•''-

IWW6-30 UBa6-00l89SEP72 C SI8-a8xU0a5-38 N SI8-49/U0a5-30 HSS 5

W0a5-30l U045-08IR SUN ELa& HZ060 188-0666-R-I-N-D-2L NRSR ERTS E-10^8-12321-5 02

Fig. IV.18 - ERTS image E-1048-12.321 - Channel 5 - Tris Marias Dam Region

- 171 -

would permit a relationship between the water collected

-by a hydrographic basin and the precipitation.

2 - A strong and frightening erosion could be observed

1n dam vicinities due to the improper deforestation

and to a bad use of the soil. The strong erosion in the

Paraopeba and Sao Francisco Rivers Basins is causing

the silting up of one of the headwaters of the Tres

— Marias Dam, clearly seen in channel 5.

A protection program, after the field checking, will

be carried out because the above mentioned phenomena

is diminishing the dam storage capacity.

IV.2.10.3"- Conclusions

1 - A relatively rapid analysis of the four MSS images

channels permitted the preparation of a vegetation

mapping in great ecological groups and the indications

of erosion and soil moisture content conditions. Such

map (Fig. IV.19) was obtained with minimal efforts,in

small periods of study, and with costs highly compen-

satory, when compared with the necessary effort to the

execution of a same kind of work using traditional

methods.

2 - It is possible to infer, from observable image surface

characteristics, areas with greater agricultural

potential.

- 172 -

[XN\- ; ^\ • v ' \- s -\\\x^ \0\o^•-s -\\\'\N--X-.S^^

V'> \> - . \ \ \x\-.\>v\>Nxi • •- , N \ sv"-* - - f

•^N^y-'

&

mm wr^ y£F& -s\ .^ /TV-> ,^* ./*v '«• ^C * '' »"^ ' •' r * ' '• • - ' -^P'/iW ^ : v v ; e ' : 2 ^ - > ' V V -f W-i^:^V:::^^v:v:/ fV-/:/O(V;:;faAV^;;

fe®m%^&V$\ ' . •• « :/v \\V »

^T^

I t

I*- * ^1*

« • 1 «

FOREST/"CERRADAO"

"CERRADAO"/"CERRADO"

"CERRADO"

SPARSE "CERRADO"/FIELDS

ERODED SOIL OR SOILPREPARED FOR CULTIVATION

"CERRADO" IH UNDULATEDRELIEF

Fig. IV.19 - Thematic map of the "Tres Marias"Dam Region.

- 173 -

3 - It is possible to identify areas with agricultural

activities and estimates the occupation intensity.

IV.2.11 - Conclusions

Even through a preliminary analysis it is possible to

conclude that the ERTS-1 images are useful for the following soil

resources survey:

1 - Natural vegetation mapping in great phytosociologiial

comunities. That mapping could be obtained with

little effort, in a record time, and with costs

highly compensatory when compared with convention-

al methods.

2 - Mapping and calculation of the areas occupied with

, spots related to remainder forests.

3 - Identification of areas with agricultural activities

and its intensity occupation estimation, as well

as the identification of crops in particular cases.

4 - Delineation of the drainage system, soil permeability

indication, its moisture degree and eroded areas.

5 - Pedologic survey, although, through a superficial

analysis: these images demonstrate great potentiality

principally related to schematical surveys or even

- 174 -

reconnaissance level on the 1:500,000 scale. In

this case it is good to keep in mind the necessity

of a field check.

6 - Channels 5 and 7 were the ones which best provided

information for the observation of the resources

here analyzed. However for particular features

analysis it is important a four channel analysis.

IV.3 - MAPPING OF NATURAL VEGETATION DISTRIBUTION OVER CENTRAL

EASTERN BRAZIL FROM DATA OBTAINED BY ERTS-1

IV.3.1 - INTRODUCTION

Multispectral information specially about vegetation types,

obtained by ERTS-1 through its MSS sub-system, fits very well with the

essential need of a large country like Brazil, to develop a natural

resource program. Natural vegetation units of central-eastern Brazil (an2

area of 622.616 km ) are mapped. An interpretation key has been established

and presented over five training areas, located in different vegetation types 1

Plant morphology, physiognomy and geomorphological characteristics are

considered to identify each habitat. A special legend developed by our

group illustrating different vegetation habitats of Brazil, is presented.

- 175 -

These habitats have been verified by prominent botanist and plant-ecolo-

gists. Boundaries for the area under study are sharp between: Atlantic,

mixed and seasonal forests (spiny and sclerophyllus), Brazilian savannah

(cerrado, campo limpo) and grasses (campo). The results are presented in

the form of a vegetation map which shows the location of vegetation types

in terms understandable to the users: governmental planning authorities,

public services agencies, investment and special research institutes.

Natural vegetation results from the struggle for existence

between,-various plant species and wildlife factors, in order to reach a

natural balance. Under protected conditions, this balance is true nature,

sinte the influence of man can be neglected. The ideal way to develop a

natural resources management program, without destroying the resources,

is to collect enough information in order to understand the working of

the environment.

Therefore the main objective of this topic is the

identification of the vegetation habitats through ERTS-1, MSS imagery

interpretation. These vegetation habitats can be applied by the end users

to develop their specific programs for planning, checking and detailed

specialized researches.

- 176 -t

IV.3.2 - Studied Area and Methodology

IY.3.2.1 - Area

The area of study under consideration includes the two states,o

Minas Gerais and EspTrito Santo; a total area of 622.616 km (Fig.IV.20).This

area has been chosen because of the wide variations in its environmental

parameters. Following Thornthwaite climatology classification Ql] , the area

varies from super humid climate in the eastern coastal part - covered by rainy

forest - to semiarid climate on the northeastern part - covered by seasonal

deciduous spiny vegetation (Caatinga). The topography also varies from low

plan topography - at sea level, to mountains higher than 1500 meters.Between

these two extremes, there are different geomorphological formations.

From a space sensing point of view, this area has been

chosen, with its many variations, to test the capacity of the ERTS's

multispectral scanner (MSS) in offering the basic information necessary to

identify different vegetation habitats. The integration of the information

from the four MSS bands provide a better image for identifying the target

and its environment than would be obtained from each individual bands.

The area under study is one of the most important regions

in the Brazil's development'plan because of its water resources potential,

forest resources, hydroelectrical and mineral resources. The lack of

sufficient regional information is one of the major limiting factors in

the plan's progress. For this reason INPE selected three test sites of

- 177 -

different natural formations to develop its remote sensing study. -

IV.3.2.2 - Methodology

The data obtained to develop this survey is from the MSS

imagery scale 1:1,000,000. By looking at any plant reflectance curve,

obtained in the laboratory by spectrophotometer or in the field by spectro-

radiometer (such as Fig.IV.21) [2], and placing the four MSS bands on the

curve, we can see that there are different information in reflectance units.

Different plants reflect incoming radiation in different wavelength depending

on the cell wall, shape and its contents [3,4,5] . This means the plant

reflectance depends on the plant itself, rather than on the physiological and

environmental conditions. The existence of chlorophyll 3 causes high absorption

in channel 4 (0.5 - 0.6y ) and 5 (0.6 - 0.7y ). The high reflectance in

channel 6 (0.7 - 0.8u) and 7 (0.8 - l.ly) is a result of the cell wall

reflectance, its state or maturity and physiological activities. The larger

the leaf area index, arid the denser the vegetation canopy cover per unit

area, the higher the reflectance would be [4 ]. Figure IV.21 can also be

taken as an example of air dry soil reflectance. Soil reflects relatively

higher in channel four compared to vegetation. In channel five the situation

is nearly the same by comparing the total area under each curve and

considering it as the reflected energy value.This might erroneously lead

one to assume that both channel can provide the same information. However

channel four is suffering atmospheric interference and appears hazy. This

filters the information and reduces its quantity on the frame and enables

us to identify the low reflected areas as vegetation, in spite of the low

- 178 -t

contrast vegetation canopy cover density data can be obtained from channel

five, because of the relatively high soil reflectance and low vegetation

reflectance. In the near infrared wavelength, the phenomena works in an

inverted way. Canopy cover reflects more than soil. Channel 6 (0.7-0.8u)

provides information between the end of the visible and the beginning of

the infrared. Bodies of water and plants growing under moisture stress

looking darker in this band. The difference in the reflectance between

soils and canopy cover in this band is greater than in the other channels.

Close, dense canopy cover reflects more than an open shallow one.

By comparing the reflectance on each band-in view of the

previous considerations, and having sufficient background about vegetation

cover, its physiography, leaf morphology, soil properties and the geomorphoj_

ogical feature of its habitat - a precise identification can be made with

the MSS imagery. In order to do this work it was necessary to develop a

legend which defines each vegetation class in an ecological manner. There

are many published specialized work about each Brazilian vegetation type and

their environmental parameters [6,7,8,9,10,11,12] . Almost all these work

have been made in small areas because of the limited transportation and

technology [13] . Very few researchers have used black and white aerial

photography. Other research has been made on the ground, only in accessable

areas. Therefore new technology - such as the ERTS-1, MSS - is necessary in

order to undertake to map the vegetation of all Brazil. It has also been

necessary to develop a new legend for identifying these vegetation habitats

in terms useful to the end users.

- 179 -*

IV.3.2.2.1 - Legend

The purpose of this legend is to identify the different

vegetation habitat by interpreting the reflected energy units in ecological

terms.

Specialists in different field can use the legend to know

the biological components, climate, geomorphology, wildlife and any other

environmental properties of each habitat. The legend has been developed to

obtain the maximum possible quantity of information offered by the MSS

imagery in order to define each habitat. Four parameters were considered;

climate, plant physiognomy, plant morphology and geomorphological features.

The legend is divided into three physiognomy types (Table B/.l).Forest; de-

fined as tree vegetation canopy completely covering the ground unit. In a

remote sensing interpretation, this means that the all reflectance belongs

to vegetation and there are no other ground reflected targets. Cerrado; is

another vegetation type containing medium high trees, not covering the

ground unit completely, but leaving spaces which vary according to its

density, under the trees there is grass cover. The third class is campos, which

is non-tree vegetatiorifc'anopy, covering the ground completely or partially.N.

The legend is"further divided into three climatic categories;

Rain (pluvial) forest, which is the super humid.region (Thornthwaite

classification) [l] , having undefined rainy seasons. Seasonal vegetation,

occurs where a dry winter and a rany summer are well defined. Between these

two vegetation types, a mixed forest can be identified which is forest

- 180 -

composition vegetation, evergreen because of high precipitation index which

varies slightly each month. The forest is evergreen, loosing the old leaves,

but at the same time forming new ones through the year - not in a specific

season. This forest occurs over mountainous topography where the temperature

is relatively low. Sometime on higher altitude and toward to the south of

the country, it appears mixed with coniferous, Parana pine (Araucaria

braziliensis) [14] , up till the pure stand of (Araucaria) in the south.

Thus, we divided the habitat (1.2.1) of the mixed forest into; 1.2.1.1-without

Araucaria and 1.2.1.2 with Araucaria.

Other two habitats can be found in the mixed evergreen

forests, which are not on the mountains. These are: gallery forest (1.2.2),

which results from sediments in the depressions or valleys with high organic

matter and moisture content soils. Swamp forest is the other habitat appears

with the mixed forests.

Seasonal forest includes three habitats in which soils - as

the result of geological and geomorphological formation - are the main

limiting factors. Climatic characteristics generally remain in each habitat.

Caatinga (1.3.1) is a deciduous spiny forest loosing its leaves completely

during the semi-arid season, this reducing the leaf area index which is

in contact with the dry air, -in turn reducing its evapotranspiration.This'_ *

habitat is found in the depressions over stony clayey soils. The second

habitat is the forestic stage of cerrado - Cerradao. It is sclerophyllous,

semi-deciduous forest on the brazilian plateau deep well drained mainlyo

latossol soils which suffer mineral nutrients deficiency and aluminium

- 181 -

excess. Plants in the cerrado adapt their leaves to'this environment by

having hairy leathery leaves which create a static layer of air between

the tiny hairs and isolate the leaf surface from the dry air. Cerrado

plants do not suffer soil moisture shortage, but only mineral. [6,7,8,9,10].

The third seasonal forestic habitat is what would be called in Brazil,

"Mata Seca" (1.3.3). It is deciduous to semi-deciduous mesophytic forest

[12] . It can be found on the Brazilian plateau over rich soils.

The second class in the legend is Cerrado (2), The word

"Cerrado" is used in two senses in Brazil: In a wide sense, it is a re-

gional, large-scale vegetation type which forms both a vegetation and

forestic province. In a strict sense, it refers to a particular range of

structural forms of this vegetation [12] . In this paper cerrado (2) is

used in a large-scale sense, to indicate the non-forestic, but seasonal

sclerophylous vegetation forms. Then it is divided into two habitats. The

first is cerrado in the strict sense (2.1.1), of medium high on a continuous

graminea grass canopy. The second is campo limpo (2.1.2), is the grassland

forms of the cerrado, where is almost no trees in the habitat. This

indicates shallower soil than in the cerrado, found mainly on the slight

siopes.Cerrado is sometimes referred to as "savana" or "savana-like" it is

in fact quite different.Eiten [12] reported: Cerrado is different than savanna

because it is an upland vegetation, which is never inundated and with

excellent internal drainage of its soil, so that soil does not remain

waterlogged for long periods after rain.

The third class in the legend is the campos. This is

- 182 -4

herbaceous vegetation which corresponds to the first class, the forestic

one. It is divided into four habitats. The mountainous (3.1) which

corresponds to the mixed evergreen to semi-deciduous forest of the

mountains (12.1) and its two habitats. The swampy (3.2), corresponds to

the (1.2.3) and the galery campos corresponds to (1.2.2). It is a vegetation

of sediment inter-mountainous depressions which are now mainly in

agriculture use. The fourth habitat is the vegetation of the coastal dunes

(3.4).

IV.3.2.2.2 - Interpretation Key

Five areas have been selected in the region under study to

see how each habitat appears in the four MSS imagery channels. The selected

areas are located in both the pure ecosystem and the transitional zones.

Three of them are INPE's test sites. These were flown over by INPE's own

aircraft mounted sensors for calibration and detailed studies. The other

two areas have been selected on the basis of recent published work which

were based on black and white photointerpretation. Table IV.2 shows the

materials used and the information obtained. The work has been done using

printed MSS imagery in the scale of 1:1,000,000 for the four channels.

Generally speaking, although channel four is always hazy, it can be used

to distinguish the vegetation*covered areas from non vegetation, because

of the relatively low reflectance of all plants in this band (Fig.-IV.21).

Channel four provides very little information.

- 183 -

The comparison between the different areas has to be made

on the same frame. It can then be extrapolated to the other frames in a

relative sense, because the sun elevation angle is a changing variable.

The angle changes during the ERTS tracking and also varies from one orbit

to another, changing the characters of the incoming radiation, in turn,

changing the reflected radiation from the terrestrial targets. The leaf

area index has been examined through infrared imagery. The greater the

leaf area index is,the higher is the reflected energy per unit area

[3,4,5,13] . In this way, we distinguished the forestic cerrado form

(Cerradao), from the vegetation form. Channel five offers more information

in the visible band and thus geomorphological information can be obtained.

Vegetation canopy density can also be obtained from this channel in an

inverted value from channels 6 and 7.

Based on the previous considerations, the habitats under

study have been characterized as follows:

- Rainy evergreen, broad leaved Atlantic forest (1.1.1.2)

These are low reflected areas in the visible bands 4 and 5.

But they are very high reflected in the infrared because they remain ever-

green throughout the year. These are very dense forests, demonstranting

relatively more reflected radiation than the other less dense vegetation

types. The eastern boundary of this habitat could also be obtained clearly

from channels 6 and 7. The habitat extends over nearly plan area, with

slight variations, extending tov/ard the. interior of the country until the

- 184 -

base of the mountains generally coinciding with the contour line of 500

meters.

- Mixed, evergreen to semi-deciduous, mountainous forest

without Araucaria (1.2.1.1)

This habitat can be distinguished as a forest, following

the previous criterion. Topography features obtained through channel five

show the drainage system, indicating the mountainous topography. Where the

forest is growing on relatively low lands, it appears darker than upland

forest in channel 6. Itmight be that this is because of the soil moisture

excess on the depression, which appear in a different image tone than the

gallery forest.

*

- Mixed evergreen to semi-deciduous mountainous forest

with Araucaria: (1.2.1.2)

This habitat is different from the previous habitat in two

aspects.The ecological difference is the misture of evergreen broad-leaved forest

with the coniferous Parana Pine (Araucaria braziliensis). The second aspect

is the climate, which permits the appearance of the coniferous species. This•

is because of the altitude-higher than 1500 meter-which presents low

temperature and high humidity indexes.

The*last two habitats are an ecoton between the high latitude

coniferous forest and the broad-leaved deciduous forest [14]

- 185 -

- Mixed, evergreen, gallery forest (1.1.2)

These are evergreen forests and can be found with mixed

forest, seasonal forests, or even campos. These have low reflectance in

the visible bands and high reflectance in the near infrared bands. They

always occur in valleys, where the soil is rich in organic matter and

soil moisture content, creating a special micro-climate. Forest reflectance

is not only feature to identify this habitat. The ground feature of

the drainage pattern can also indicate the valley direction, the sediment

resources and the valley mouth. The forest density is always lesser on the

top of the gallery than on its mouth following the soil depth direction.

- Seasonal, deciduous spiny forest (Caatinga 1.3.1)

This is a forestry habitat of thorny leaf plant, which adapt

their leaves to be able to live under semi-arid climate. During the dry

season plants loose their leaves and have a low leaf area index. These two

properties present low reflectance in the infrared bands. The trees usually

cover the soil, which do . not permit detectance of high soil reflectance

in the visible bands. For this reason it appears dark in both visible and

near infrared bands. Caatinga habitat can be also distinguished, with high*

precision, through geomorphological feature. It is always found on the

(depressions, inter-mountains or inter-plateaus, on stony clayey and poor

drained soils. This information could be obtained from channel 5.

- 186 -

- Semi-deciduous, sclerophyllous forest (Cerradao 1-.3.2)

In this habitat the forest plants adapt to the dry season

in a different way, leaves are hairy leathery sclerophyllous [6,7,8,9,10] .

The importance of this characterisitc is that the cell wall is harder and

the cuticle is thicker, causing relatively high reflectance in both

visible and near infrared bands; especially in channel 6 due to the well

drained soils.

- Mixed, semi-deciduous to deciduous forest (1.3.3)

It grow over richer soils than cerradao. It does not loose

its leaves completely during the dry season and does not suffer from lack

of nutrients or moisture content. For this reason, the leaves are thiner

than cerradao, reflect less in the visible bands. Moreover it can suffer

water stress, and this can be seen on channel 6.

- Seasonal, sclerophyllous cerrado (2.1.1)

This is the habitat which follows the same characteristics

of the cerradao, but with higher reflectance in the visible bands.because of

the lesser vegetation canopy density, permitting high soil reflectance to be

detected, in some percent, by the sensors. Channel five shows here the soil

condition and the geomorphological level of the plateau.

- 187 -

- Campo limpo (2.1.2)

This is the grassland form of the cerrado which indicates

shallower soils than the previous habitat (2.1.1). It has a higher reflectance

in the visible bands because of the poor canopy cover, which enables the

high soil reflectance to be detected. There is less reflected energy in the

near infrared because of the small leaf area index. This habitat has a

shallow soil frequently occuring on the slight slopes on the boundary of

the Brazilian plateau. The drainage patterns obtained from channel five

indicate the slope degree and direction.

- Campo (3.0)

» j

This class is a short grass type. It has high reflectance

on channel four and five due to the high soil reflectance and relatively

low leaf reflectance. This class indicates four habitats with respect to

different ground features:

- Campos of the mountains (3.1)

This is the grassy stage of the nixed evergreen to semi-

deciduous mountains (1.2.1). It has shallower eroded soils, causing higher

reflectance in the visible bands in comparison with the mountainous forest.

At the same time there is low infrared reflectance because of these eroded^

and lesser vegetation density canopy cover.

- 188 -»

- Swampy campos (3.2)

This is a vegetation canopy habitat over shallow water

table level. It has low reflectance in both visible and near infrared

bands, because of the low reflectance of the wet soils in the visible

bands, the low vegetation reflectance in the near .infrared bands, and

water absorption. The absence of drainage patterns indicate its

boundary. If this swamp is near the Atlantic coast, then it would be salt

marsh with saline water table. If it is in the interior, then it would be

fresh water table swamp. Of course plant communities are different in

each case.

- Gallery campos (3.3)

This is wide valley habitat between the forest covered

mountain. They are now developed for agriculture tuse and are distinguished

geomorphologically, and in any channel by the maun made patterns.

-.Coastal dunes campos (3.4)

The dunes themselves are high reflectance targets. If there

is vegetation growing on the-dunes, then dark patches will be detected in

channel five. It is easy to draw the boundary of this habitat through both

channel six and seven, because it is contrasted with neighbour habitats;

the ocean to the east and the swamp or Atlantic forest to the west.

- 189 -i

This is basically the way we interpreted the ERTS-1, MSS

imagery to develop the map.

IV.3.3 - RESULTS AND CONCLUSIONS

1 - The map enclosed (Fig. IV.22 - "Mapa de Vegetagao Natural"

Natural Vegetation Map) is the final output of this work,

having transferred the reflected energy units into a

common understandable language. It shows the following

vegetation distribution: The Atlantic coast in the eastern

boundary, with the presence of the sand dunes in a dis-

continuous strip, broken sometimes by swamps or directly

by the Atlantic forest (1.1.1.2). Moving toward the west,

the mixed evergreen forest (1.2) and the mountains campos

(3.1) appear on the mountains topography. Upward, toward

the west, the sclerophyllous forestic vegetation can be

found. If there are deep soils, then cerradao (1.3.2) will

be found.Shallower soils would be covered by the cerrado

(2.1.1). On the slight slopes the vegetation cover is campo

limpo (2.1.2). The sharp slopes do not permit sclerophyllous

vegetation to grow, this is campo (3.1) on the slopes and

mixed fore-st frequently present. In the northern part of

the region, on the inter-mountain depressions, the canopy

is piny forest-Caatinga (1.3.1)., The mixed semi-deciduous

forest"mata seca"(1.33), is also a plateau vegetation, but

is found over richer soils in some shallow wide depressions,

especially near rivers.

- 190 -

2 - It is therefore possible to get the basic ecological

information, presented on the map, through the MSS

imagery. Geographical location of the mixed forest that

1t is a transitional geomorphological zone between the

Atlantic forest (1.1.1.2) and the cerrado (2). This agrees

with the Brazilian geomorphological explanation of

AB'SABER [15 ].'

3 - The same map proves that the mixed forest (12.1.1) is an

ecoton between the evergreen and the coniferous forest

(12.12), which agrees with the ecology point of view of

EVER [14]:

IV.3.4 - DISCUSSION

We have classified the end users of remote sensing technology

into: Special research institutes, governmental planning authorities, public

service agencies, and investment and development organizations. The MSS

imagery has been shown to offer enough multispectral data to be used to

obtain basic ecological information using a well oriented legend and

interpretation key.

*

The relation between the vegetation cover and its environment

also provides information for the other specialists, who are concerned with

natural resource investment and management. The fact that a large area can

be scanned in a short period of time is an important advantage in a multi -

spectral interpretation. It can also provide information for a large area

- 191 -

under the same physical conditions, such as: sun elevation angle,air

temperature and humidity which may change the reflectance of the vegetation

to adapt its physiological function to the environment. Care must be taken

to compare each frame independently since different frames are scanned

under different atmospheric conditions.

The special research institutes can work inside each habitat

to get more detailed data and compare the different locations of the same

habitat to better understand the different phenomena.

Governmental planning authorities can use the habitats as

indicator to the land potentiality. On combination with other socio-economic

factors, knowledge of land potentiality will enable authorities to establish

well planned programs.

Public services agencies can, with this knowledge investigate

whether or not land is being used most efficiently.

Investment and development organizations can establish

priorities and a credit system based on the knowledge of how to best use

the land.

*

We have then demonstrated the capacity of multispectral

scanner in providing the basic information necessary to develop a Brazilian

natural resources program.

- 192 -

IV.3.5 - References

[l J Departamento de Meteorologia, Atlas climatologico do Brasil, Rio de

Janeiro, Brasil,1969.

[2] R.M. HQFFER, ADP of Multispectral Scanner Data for Land Use Mapping.

-.. Information Note 080372. LARS - Purdue University, West

Lafayette. Indiana,U.S.A., 1972

[3] W.E. LOOMIS, Ecology 46, 182, (1965).

[4] W.A. ALLEN, and A. J. RICHANDSON. "Interaction of Light with a Plant

Canopy; Proceeding of the Fifth Symposium on Remote

Sensing of Environment, University of Michigan, Ann Arbor,

Michigan, P.219, (1968).

[5] D.M. GATES, et al. Applied Optics, £. 1: 11 - 20 (1965).

[iej M.G. FERRI, Contribuigao ao Conhecimento da Ecologia do Cerrado e da

Caatinga. Estudo Comparative Balango D'agua de sua Vegetagao.

Bol. Filgs. Univ. Sao Paulo, 195, 1-170 (1955).

j_7j M.G. FERRI, Problems of Water Relations of Some Brazilian Vegetation

Types with Special Consideration of the Concepts of

Xeromorphy and Xerophytism:. In plant-water relationships in

arid and semi-arid conditions, pp. 191-197,UNESCO, Madrid

Symposium (24 - 30 Sept. 1959)

I 8 I K. ARENS, "Consideragao s6bre as Causas do Xeromorfismo Foliar".

Bol.Fac.Filos.Univ. Sao Paulo, 224, Bot.l_5: 25 56. (1958).

[9] K. ARENS, "0 Cerrado como Vegetacao 01igotrofica" bid. pp.59-77 (1958).

- 193 -

[_10J K. ARENS, "As Plantas Lenhosas dos Campos Cerrados como Flora,Adaptada

as Deficiencies Minerals no Solo".Em simposio sobre o cerra-

do, pp. 255-303. Editora. Univ. Sao Paulo (1963).

[ill L.G. AZEVEDO, "Tipos Eco-Fisionomicos da Vegetagao da Regiao de Ja-

nuaria (MG). Em II Simposio sobre o Cerrado. Anais da ACA.

Br. de Ciencias, Rio de Janeiro, pp: 39-57 (1966).

[l2J E. EITEN, "THe Cerrado Vegetation of Brazil, In the Botanical review,

138, 2, pp. 201-341 (1972).

[l3j L.H. ALLEN, and K.W. Brown, "Shortage Radiation in Corn Crop", Agr. J.

57, pp. 575-580, (1965).

[l4J S.R.EYER, "Vegetation and Soils - A World Picture". Second edition

London, England (1968).

[iSJ A.N.AB'SABER, "A Organizagao Natural das Paisagens inter e sub-tropicais

Brasileiras. Ill Simposio sobre Cerrado. Editora da USP

Sao Paulo (1971).

- 194 -

TABLE IV.I - LEGEND

1.0 - Forests

1.1 - Rain1.1.1 - Evergreen, Broad!eaved1.1.1.1 - Amazonic: a - Terra firme (UP LAND)

b - Varzeac - Igapo (LOW LAND)

1.1.1.2 - Atlantic1.1.1.3 - Palm dominated

1.2 - Mixed1.2.1 - Mountain1.2.1.1 - Without Araucaria1.2.1.2 - With Araucaria1.2.2 - Gallery1.2.3 - Swamp1.3 - Seasonal1.3.1 - Deciduous, Spiny (CAATINGA)1.3.2 - Semideciduous, Sclerophyllous (CERRADfiO)1.3.3 - Mixed, Semideciduous to deciduous (MATA SECA)

2.0 - Cerrado

,--*2.1 - Seasonal, Sclerophyllous2.1.1 - Cerrado

v

2.1.2 - Campo limpo«

3.0 - Campos of:

3.1 - Mountains3.2 - Swamps3.3 - Gallery3.4 - Coastal Dunes

- 195 -

IIIIIIIII

01

ra•MtoO)

to<aa>ro

ro

O)

CM•

i—I

a*-Qra

ro*Ja>a.ai

a>JC.4J

oUIzt— 1•Xt—CO CO0 I—

sr i—O >-•I— 1 CO!_• ^^< f T*

^»CUOU.Zfe>-4

OUI

^3

CO^J

<c1— «Q£UI1—

^^

X

<cU. UIO «v

ÛIs: uir T"

Z I—

»*^X^»^ ^^ i i^

CO CM f— CM• • • •

CO CO i — i —• • • •

c— t— CM CM

OCL

10 EO O -r-OMro O •—co -a TJ

ro ra O(0 1- i. Q.-(-> 1_ i- ErO (U (1) <O:E o <_> o

cor-

«3- CM CMvo r»« r^ 2 ?1

Oo

-oM

O*o• •

rv

a><ouC/)

c" O

to •«-0 •*->•M to rao ai E.C CD CQ. ra O

E «*-O> -i- C

• ^™

-C ro -C2 c i »

4-> 3• o u t -C O -4->n> a.

to "oê i- c0 4J 3ra •— o

•— 3 J-oo s: o

a>croIMrO

^

to

o — *C ro r—

IO i. CO•4-> ro *-*C U33 MO ro CE i- t-

M ^ j^C 4->OJ -M Cai 3 3S- 0 0ens: E1- -4-> x ^- oO)

+J i— 10-o to .0

COt*»

CM CMf^ r »»a> cn

oo0»inCM• *

f™"

ai*fOU(y)

C«. o

«/) t-O -4->

••-> ro0 Ef^ t

Q. OCâi c

^> *r**r~J= JC2 ••-«

3•a t.c ••-»ro

-o-i£ CU 3ro O

•— 1-CQ CS

rocnc

•r—•Mroa.

^ "^CM

•r»

•— » CMr~ *^^

•

<— Oex

CM E

|-»

O-o oro CXX_ EL. roai <_>«_, ^^

cr> cn ^~vo ô r^cn cr> cn

0 0 0o o o0 0 C3

in 'in oCM .— f—• • • • • •r~ r** r*~

O) O)

rO <OU Uvt vt

«k «%tO (/)o o^* «J

o o.C J=0. 0.

CI "X34J 0)•^ s--C roS »-

-a cC >— ira

t.J* 3U Oro t—

<— OCO (_>

a>(U•^roM-<a~"

^ "V

--» -. CMCO i—

• • ^"— 'CO r—

• CMr— CM « — '«^ » ^^

Oa.

ra Eo •«-ai o _jco -o

rO Oro S. CX•M J- Era 0) IO. <_> CJ

^"VOcn

oo0

«k

o o0 Oo •—CD fCDCM U

r— «0

(U Or- EraU O

oCX CXroE T3

atc •—O t—t- O•M l-ra 4->

OJ OCD O0) C

• (/)roC

^3U_

*•" X™ *»

r*^ r~• •

CO r-

r— CM^MX **-*

too> oc -a•i- ro4*^ &~ro J-ra (U

<_> CJ

IO ID10 t^cn cn

0oo

M

0 00 0O LO

^ i—cor— Q>• • r**r- ro

U

0c •—0 0•r- -C4-» CXro i-J-> O

t/1

-M10

uiQ.

airatoraOJ

to01a>

10t.

<D

Of4->O

- 196 -

40^ EQUATOR

- 10

^•1 MAPPED AREA

Fig. IV.20 - Lpcation of the studied area.

- 197 -

L 13NNVHO

9 13NNVH*9*

fi 13NNW9'

* 13NNVH9

otf>

o»o

Ocu

(OccUit-UJ

o(Co2

o

UJI

ro

4->U«*- t/» i_

t- cuO) C7>> cu<c >

CM

O

5 's-o

0>

- 199 -

IV.4 - ESTIMATION OF PASTURE PROJECTS USING ERTS-1 IMAGES


\

This work is executed in cooperation with the Superintendence

for the Amazon Region Development - SUDAM, It is a planning and coordinationi

Federal agency for the legal Amazon Region. In the private field one of its

objective is the analysis, the approval and supervision of the Pasture

Projects introduced in this region through fiscal incentive given by the

Brazilian Government.

The research under study aims to obtain, .using ERTS-1 images

a rational and productive methodology to the control and following of the„'•.

implanted Pasture Projects evolution, through specific studies of the

following objectives:

- To,survey areas under dense forest, "cerrado" vegetation

areas and flat lands.

- To delineate different drainage patterns.

- To determine the road system and estimate its extension.

- To control the deforestation using the ERTS-1 images

repetitivity.

- To evaluate areas with pasture.

NOT

- 200 -


To attain the objective of this work, it was selected a

test site (described below) to permit the evaluation of the MSS/ERTS-1

images capability for this kind of study.

IV.4.2.1 - Test Site Selection

The area under study is located at the Northeast of the .''ato

Grosso State, comprising an area between the parallels 10°00' and 16°00'

South and meridians 51°00' and 54°00' West, embracing parts of the Xingu

and Araguaia Rivers Valleys and the Banana! Fluvial Island (about 300,000

km2).

This area was chosen because is suffering an accelerated changei

due to the great number of farms (cattle breeding and cropping) already

established.

There is not a recent photographic coverture of this area,

which is out of the designated area for the RADAM Project (natural resources

surveys through the SLAR system in the Amazon Region). Besides this, the

area is under the jurisdiction referred to the 50% of forestal reserve as

established by the law n9 4771, dated September 15,19.65 (Forestry Code) in

its article 16, letter b.

- 201 -

IV.4.2.2 - Approach

After the preparation of basic maps (vegetation.drainage

and road system), the properties delimitation should be done over this

area mosaic (ERTS images on the scale 1:1,000,000), with the help of data

of the SUDAM Office* general information as welU as natural features.I

After this phase, periodical evaluations should be done

at 6 months interval of this region through the ERTS-1 images mosaic

interpretation on the scale 1:1,000,000. The mosaic analysis will permit

the.selection of frames, where visible changes appeared in the region for

image enlargement to the scale 1:250,000 with a better precision and with

pasture areas measurements.

Periodically, after mosaics interpretation, a field checking

should be done to improve the limits of each Pasture Project as well as to

verify the accuracy of areas measurements on images and also provide

information such as grassland, planting orientation, carrying capacity,etc.

For illustration purposes two ER1TS images in,channel 5, taken

at different data, are presented here in figures IV.23 and IV.24. They are

areas comprising this test site under study greatly justifying the viability

of this project in that special region. It should be observed with what

contrasts degree these pastures areas could be separated from natural forests

and from the accentuated deforestation this regriion is suffering, through

the comparison of the two different gathering data of the ERTS images.

I- 202 -

U05I-30I U05I-00I U050-30I37NOV72 C SI I-33^1051-05 N SI I-35/U050-59 OSS 5 R SUN EL58 RZI02 188-la89-G-I-M-D-2L MRSH ERTS E-1 107-12593-5 01

I

I

I

Fig. IV.23 - ERTS image E-1107-12.593 - Channel 5.

- 203 -

-1**' ' "'''• -." . 'S>j '. ": - - ' '" ' I * *V» •'t r.-*•-, __ y v- _-.^- . fc*;" '<'i- • *; '

* "•- - P-A. i --•W-3'"'!' ' .-r-'" , "'X ' »•-.

:t.f " l--::.." -:--- '-** • 1 -, -••? -//-W t- ». , • r

-v ,-W."?>fe

" ^

Fig. IV.24 - ERTS image E-1377-12.584 - Channel 5.

- 204 -

IV.4.3 - Preliminary Results

The final report related to this topic is expected to be

ready by the end of 1975.

Up to now two basic maps (enclosed - Figs. IV.25 and IV.26)

were done based on the ERTS images. The first one shows the drainage

system, the road system and cities and the second one shows the natural

vegetation map as well as the delimitation of projects already installed.

The mapped units as well as their patterns appearing in the ERTS images

could be seen as follows:

IV.4.3.1 - Preliminary Interpretation

The MSS channels 5 and 7 were used.

IV.4.3.1.1 - Channel 7

It was used to have the hydrography overlayed.

o

- The riverbeds are not covered by vegetation presenting

themselvesJn dark gray tones, as well as lakes and

lagoons.

- Soils with sparse vegetation or bare soil which present

a high degree of moisture, appear in dark gray tone as

for instance the lowlands between the Xingu and Mortes

Rivers as well as the Bananal Fluvial Island.

- 207 -

- The small water bodies, circled by dense vegetation,

shown ti the image lighter grey tones, and as this

vegetation enters in a degradation phase, passing to the

arboreous and shrubby formations, the tonality tend to

go to the darker grey tones.

- Through the ERTS images E-l305-4501 and 1305-2501 (IE and

2E) it is possible to note sparse darker grey tones spots

(almost black) due to their localization always close to

the water bodies. In areas which the surface drainage

pattern is highly dense, it is supposed they are denser

"cerrado" vegetation areas with a high content of moisture

in the surface.

- Roads and cities appear badly in this channel.

- A comparison was done between the map done using the ERTS

images (Fig. IV.26) and the USAF navigation map on the

scale 1:1,000,000.

Superimposing the .two above mentioned maps the main water-

bodies are with their forms and localizations correct, excluding the Mortes

River, at the South of the area, which has its form strongly different from

that one presented by the ERTS images.*

Another observation done is that some lagoons showed in the

image do not appear in the USAF and IBGE charts. For instance a lagoon in

the margins of a Culuene River tributary, seen in the image.

f3HCSDING PAGE BLANK NOT FILMED

- 208 -

IV.4.3.1.2 - Channel 5

Without any previous field work in this area, our main

objective was to map the main forestry formations, to delimit artificial

pastures, to localize road systems, cities and small towns of the region.

1 - Vegetal Coverage

Basing principally on the grey tones and through the

comparison of channels 5 and 7, it was possible to preliminary map the

following:

1.1 - Dense Natural Vegetation (M)

It approximately embraces 2/5 of the area and presents

itself in dark grey tones (almost black).

Due to the penetration facilities in this forest :

area through the road n9 BR-080 (Barra do Gargas-CachimboJ.and with the

increasing implantation number of Pasture Projects, mentioned before, it

could be deduced that this region will suffer a continuous deforestation

for artificial pasture formation, what is currently happening as could be

seen in the images.

It is known that the forest areas are preferred for the

above mentioned purpose because they generally have better soils for artificial

- 209 -*

pastures development.

1.2 - "Cerradoes" (CC)

"Cerradoes" areas are presented in intermediate grey

levels tending to the dark. In some images they were not separately mapped

because they are strongly mixed with "cerrado" vegetation.

1.3 - "Cerrado" (C)

It was tried to map three units with "cerrado" as

follows:

1.3.1 - (Cl)

It was characterized because it presents little

surface drainage and more rough texture. It is supposed to be more

permeable type of soils.

1.3.2 - (C2)

It is characterized by a dense surface drainage

system, making us suppose to be a more compact type of soil.

- 210 -

1.3.3 -

It presents a lighter gray level in channel 5.

Comparing channels 5 and 7 it is noted an area with a very sparse and

humid "cerrado" vegetation, comprised between the Araguaia and Mortes

Rivers. There exists in the area several pasture farms leading us to

suppose that these humid "cerrado" were used for natural pasture in

its great part.

1.4 - Galery Forests

They are easily mapped along the water courses.

2. Artificial Pastures

They are easily seen because of their geometrical forms,

principally when localized inside a dense vegetation. When situated in the

"cerrado" field zone, due to the lack of contrast (light gray tones) there

are not many clearings needing comparisons in the field for a better

delimitation.x

Calculations of the pasture areas will not be done due to

the inadequancy of the ERTS images scale for this kind of study.

IV. 4. 3. 1.3 - Cities

The major cities were mapped based on their radial forms

- 211 -

and their proximities to roads, but small towns and villages are

confounded with other features in the area. = :" -

IV.4.3.1.4 - Roads \ .'

They are possible to be mapped and are very clear.

IV.4.4 - Other Considerations

1. Through the ERTS image number 1304-5501 (3E), channel 5,

— it 4s noted that in the margins of Xingu, Culuene Rivers

and others, a lighter tonality appear, contrasting with

the dark gray tone of natural forest. It is inferred that

they are hydromorphic soils with little infiltration

capacity, forming extense lowlands with less dense

vegetation related to the adjacent forests. This lowland2

was planimetered, possessing about 1,185 km .

2. Some lagoons present a lighter gray tone similar to those

showed by rivers while other seem very dark. It could be

-êxplained by the great quantity of suspended material in

the first mentioned ones. It is also inferred that those

lagoons should have a direct communication system with

the river.

- 212 -

IV.5 - IDENTIFICATION OF FORESTAL COVERAGE THROUGH THE ERTS-1 IMAGES

l!


\

This work was developed in collaboration with the U.F.V.

(Federal University of Vigosa - Minas Gerais State) and the ESALQ (LuTs

de Queiroz University - Piracicaba - Sao Paulo State) .

Its main purpose was to determine the capacity of certain

remote sensing system to provide useful information to the forestry

problems solution, through studies of:

- mapping of native and artificial'forests using the

data obtained from the ERTS-1 satellite.

IV.5.2 - Materials and Methods

IV.5.2.1 - Materials

The following materials were used:

c

- ERTS image E-l388-12.205, dated August 15,1973, on the

scale 1:1,000,000.

- The same.ERTS image enlarged to the scale 1:250,000

- Black and white aerial photographies on the scale 1:30,000.

- 213 -

IV.5.2.2 - Methods

I;- Test Site

The Ipatinga Test Site was chosen due to its localization

between two large rivers: Doce and Piracicaba and related to vegetal

coverage it presents a great natural forest named the State Park of Doce

River. In this State Park there are lakes with more than 5 km of extension

and clues of old lakes where characteristic vegetation developed. Round

about this forestal park it was found great Eucalyptus plantations from

several growing ages.

This test site is located at the Minas Gerais State presenting

the following coordinates: 19°25'30" North - 42°25'00" East and 20°00'00"

South - 42°37'30" West.

- Approach

An aerial photography mosaic was done in the scale 1:30,000.

From this mosaic a basic map of the area was done containing vegetation,c

drainage, road system information, and so on. This basic map was made

from photographic interpretation on the scale 1:30,000. This map presents*

several types of vegetation verified in the photographies, with some of

these areas checked in the field.

In the ERTS image 1388-12.205, scale 1:1,000,000, a localization

- 214 -

of the test site was done and it is shown in figure IV.27. After the

localization, this part of the image was enlarged to the scale 1:250,000 -

channels 5 and 7 - which were used for interpretation.

\IV.5.3 - Results1 i

\In this image was possible to identify the following units:

old reforestation, new reforestation, natural f©rest in flat lands and

natural forest in rolling lands as shown in figure IV.28.

That interpretation was done basiing on gray tones, units

geometrical aspects, comparing with the basic Esap and with some field

knowledge. Channel 7 was used to delineate water bodies (rivers, lakes)

which present themselves with a very dark gray tone in this channel

contrasting with light gray tones of the vegetation. Channel 5 was used

to delineate the mapped units limits, considering the gray level and also

geometrical aspects of these units.

Related to these parameters, the mapped units appear in

both channels as follows: e — . - • _ - .•v

- The natural forest present itself with an homogeneous

aspect related to texture and tonality. It appears in

a darker gray tone in channel 5 and in a lighter gray tone

in channel 7.

I- 215 -

I

I

IU0<»a-80 U0a3-30l U0^3-08I S02I-00II5HUG73 C S20-05/U0<13-l2 N S20-07/U043-06 HSS 5 R SUN EL38 RZ05I !88-5a07-N-I-M-D-2L MRSR ERTS E-1388-I22B5-5 01

Fig. IV.27 - ERTS image E-1388-12.205 - shows the test site location.

ORIGINAL PAGE IBQUALITX;

- 216 -

L E G E N DSCALE_2 U_ 25.0(000

OLD-REFORESTATIONNEW REFORESTATIONNATURAL FOREST ONPLAN TOPOGRAPHY

NATURAL FOREST ONSLOPES

ROADSLAKES

Fig. IV.28 - Thematic map of the Ipatinga Test Site.

- 217 -

- The reforestation appears with an mottled aspect due to

,.r..~.r.~*-.?:-.-. .-.,,: -.-the presence of inserted small areas with definite geo-

metrical forms, caused by human activity.

- The distinction between old and new reforestation was done

considering the responses on both channels. The old reforestation presents

Itself with a dark gray tone at channel 5 and at a light gray tone at

----- channel 7, while the new reforestation presents itself with a middle gray

tone at channel 5 and a light tone at channel 7.

IV.6 - ORBITAL IMAGES UTILIZATION (ERTS-1 AND SKYLAB) FOR

PEDOLOGICAL SURVEYS


With the appearance of artificial satellites to survey

natural resources and the great potential it offers, it was decided to

demonstrate the ERTS-1 and SKYLAB capabilities to the pedologic surveying

of some type of soils which appear at the South of Mato Grosso State andc

to the North of Parana and Sao Paulo State. The viability of such a

project will bring great benefits to the Ministry of Agriculture and

to Agriculture State Secretaries for the elaboration of a Brazil

schematical soil chart.

- 218 -

IV.6.2 - Test Sites

For the realization of this project test sites were selected

in the Mato Grosso State (Dourados and Amambaf region), Sao Paulo State

(Ribeirao Preto region) and Parana State (Maringa region) as shown in

figure IV.29. In these test sites there exist a great variation of soils

and are, in their great part, occupied by an intensive agricultural2

activity, embracing an area of approximately 140,000 km.

The relevant facts for test sites selection are the agree-

ments INPE maintains with the Agricultural Secretaries of the above

mentioned States. So, researchers from these agencies will contribute

providing such ground truth information as well as soil maps in a

reconnaissance level of the above referred areas, which will be fit for

comparison works with the orbital interpretations.

In the selection of the areas some factors were taken

into consideration. Among them we considered the commom elements

of different regions and the basic substructure for the soil mapping«

project development.

These areas will be used as support points to the ground

truth missions. These missions consist in the identification of basic

- 219 -

Fig. IV.29 - Test site location.

- 220 -

parameters and their relation to laboratory measurements.

ii

IV.6.3 - Material and Methods

The project will utilize 70 mm transparencies of the MSS/

ERTS images on 4 channels and color infrared composite, as well as paper

images on the scales 1:1,000,000 and 1:500,000.

The infrared color SKYLAB transparencies and compositions

done with MSS transparencies will be used as support to preliminary inter_

-pretations of different types of soils basing on the different spectral "

responses of the areas under study.

The methodology to be used is based on bibliographical

information of regional works related to the test sites under study.

Contacts with local research institutions will be carried on.

During the ground truth missions for the gathering of

ground truth data these will be surveys of physical, temporal and spectralc

parameters of the interested target. Soil sample's collection (which will

be submitted to physical, chemical and spectral laboratories analysis)

will be related to bibliographical information and soil maps of the

areas under study. This process aims at the establishment of limits

occupied by the different types of soil through ERTS images.

- 221 - ' .

After the preliminary images interpretation related to

soil types, densitometric measures will be done of the respective

patterns and submitted to a computer program for a possible boundary

correction. v

i

Soil maps of the different test sites will be done as

final products following classical presentation patterns aiming at a

practical and immediate utilization by the users.

IV. 7,-,:STUDY OF:THE CHANGES IN THE LAND USE ASSOCIATED WITH THE

"TRANSAMAZONICA" ROAD DEVELOPMENT. USIK6 ERTS-1 IMAGES


-,-..,- . .:v^^- The "Transamazonica" road crosses regions with great climate,

soil types and vegetal variations, from the Brazilian coast to the frontier

with Peru.

e

The deforestation along this road occurs in large intensity%

caused by the human settlements in this road region. It is really important

to carry out studies to follow the ecological changes caused by the

deforestation and the consequent land use by oan associated with the

"Transamazonica" road development.

- 222 -

The ERTS images permitted a regional study of these changes

and its repetitivity will permit the follow up of the pasture projects

development during one year period.

This work is practically in its Initial phase and in

collaboration with the "Escola Superior de Agriculture LuTs de Queiroz"

(LuTs de Queiroz Agriculture University) and with the Amazon Development

Office - SUDAM.

This study will be based on the ERTS-1 images visual inter-

pretation on the scale 1:1,000,000, and occasionally image enlargements

to the scale 1:250,000 in channels 5 and 7 in snore detailed studies.

RADAR mosaics, available at scale 1:250,000, will also be utilized. The

approach to be used in this interpretation is of a multidisciplinary

character to permit an ecological estimation which should be done at every

6 months interval.


c

The main objectives of this project are:

«

- To study land use development represented by pasture,

crops and forests.

- To study erosion resultant of deforestation.

- To determine sediments source and flow as results of

this mentioned deforestation.

- 223 -

;::-,,To,.study., the "Transamazonica" road influence over the

environment.

- To estimate changes in rural and urban patterns.

IV.7.3 - Activities already done

As previously mentioned, this work is in its initial

execution phase.

Up to now a survey of all the ERTS images covering this

specific area was done.

Afterwards a preliminary study was done of all the test

sites to establish relevant areas in view of the great human occupation

as well as to good ERTS images availability to this region.

IV.8 - PASTURES EVALUATION BY ERTS IMAGERY

• «,. : (MULTISPECTRAL REMOTE SENSING APPLICATION IN RANGELAND

CAPABILITY EVALUATION FOR GRAZING)*


Animal production is still the main protein resource to

- 224 -

supply man's needs. Ecologists, botanists and economists are always

working hard to give range managers new and important information to manage

ranches in the most efficient way. Almost 90% of the livestock production

of Latin America is dependent on natural range grassland, but the mentality

of conservation of natural grazing lands is not deep-rooted (3). The

shortage of meat and milk in the Brazilian market raises the need to look

for a more efficient technology to evaluate the range resource capability

for grazing. It is vitally important to have an accurate up-to-date knowledge

of the structure, use and condition of the vast areas of the Brazilian

natural grazing land. The job is very complex and expensive, that is why

it has never been done.

_ -Remote sensing techniques offer up-to-date information and

rapid feedback on changing environmental conditions. It is also necessary

to know what season of the year, difference in characteristics can best be

detected, on .which:kind of film, and using what filters (1). We want to

demonstrate that remote sensing techniques can do this job more efficiently

with fewer specialists over a shorter time period for large areas. Multi-

spectral analysis was used to obtain information for a two-year-protected,

native pasture, by flying over the area with a multispectral cameras system

from an altitude of 500 meteVs above the ground.

- 225 -


li

The chosen area is located inside INPE's test site in

Cachoeira Paulista, 200 kilometers from Sao Paulo. The vegetation cover

is mainly a grass known in Brazil as Capim Gordura (Melinis mi nuti flora)

and very few mixed areas of Capim Coloniao (Panicum maximum). The two

main invaders are Sape (Imperata brasiliensis) and Vassourinha (Baccharis

rufescens). Taboa (Thyfa sp.), an unpalatable plant, indicates the

presence of highly humid lowlands soils, and a high water table. Some

other species appear in some scattered plots because of the human use

of this area before it was placed under protection. Napir elefante

(Panm'setum purpureum), Batatais Paspalum notatum), Bambu and some

bushes appear in the area also. During June-September, the dry season

of Sao Paulo State, the palatable species such as (Melinis) are usually

dry having lost their leaves after the flowering stage and the seed

apreading stage. The invaders such as Sape and Vassourinha stay fresh

and strong extracting their needs from soil moisture but provide a less

dense canopy cover. The dry season was chosen for the flight because of

the contrast between palatable and unpalatable plants. Laboratory

spectrophotometer analysis was done to study -the leaf reflectance along

wavelength bands, using a Perkin-Elmer 450 model, with integrated sphere.»

In the growing stage of both (Melim's) and (Imperata), the leaf reflectance

curves (Fig. IV.30), show high chlorophyll absorption around 500 and 650 rma

wavelength. But (Melinis) shows higher absorption than (Imperata).

(Imperata) has the same reflectance curve pattern as (Melim's), but it is

- 226 -

displaced about 10 mu towards the shorter wavelength, 1n the all bands,

comparing with (Melim's) curves. Using this information an estimated

leaf reflectance curve was established (Fig. IV.31). It showed that

there is high contrast between areas covered by (Imperata) and (Melinis)

in the spectral regions of 500, 600 rma and after 700 my . For this\!

reason four filters have been chosen to be used with the selected films.

(Table IV.3). The combination of filter transmissions and plant reflectance

curves are also illustrated (Fig. IV.31).

The flight took place on the 6th of July 1973, using a

Hasselblad 500 EL/70mm system from an altitude of 1500 feet. Photos at a

scale of 1:9144 were thus obtained. Four filters were used: yellow, orange,

red and Kodak Wratten 89B. The primary purpose of the yellow filter is the

correct reading of those color tones in the gray value. The orange filter

is more contrasted filter and over corrects for the blue sky, so that,

sky tones become darker than they appear to the eye and distant details

appear stronger. The red filter goes further in over correction (4). Table

IV.3 shows the flight parameters. The operation time was around 12:00

o'clock to avoid vegetation shade.

IV.8.3 - Interpretation

The data was collected from four bands, each filter

transmitting information about the ground cover. The summation of the

- 227 -

\

UJO

§

§Offoz^^(0

zUJz

UJo

Jrt

ozccUl

1u.z

^^

z

UJ2

-

ute>20>

oz

oCE .oz

««

(EUJQ.

^

ii *

1 1 t i l l 1

. » 1

i-. i

i1'_

iij1*

- iii.• i

. / i

C

COUl>cr3oUl

oUl

UlCE

CO

Ul

6ro

>

O0>

o•

OT w co 55 z^r o:-J IJ 13 JUJ UJ IU %S S 2 J

i J

i I

UJX

CD

1<

COzoCOCO

2to

a:UJ

UJo

uUJ

UJcc

OO)

oo

o<o oto

oto o

CM

UJ5HUJCO

CO

S

ooi

ro

"Ml*

- 229 -

TABLE IV.3

FLIGHT PARAMETERS

Relative Altitude: 500 meters

Date: July 6, 1973

Time: 12:15 local time

FILM

Plus-X Aerographi c 2402Estar Base SP 494

Plus-X Aerographi c 2402Estar Base SP 494

IR-Aerographic 2424Estar Base SP 494

Aerochrome IR 24443Estar Base SP 494

FILTER

Yellow

Red

89B

Orange

VELOCITY

1/250

1/250

1/250

1/250

APERTURE OF DIAPHRAMOR f/STOP

f/11

f/8

f/8

f/4

- 230 -

data obtained from the four bands made accurate and efficient interpret- :

ation possible. The yellow filter transmits reflected radiation by the

plant surfaces longer than 0.45y wavelength. The red filter eliminates

reflected radiation in wavelength shorter thara 0.575u , Thus the red

filter gives information in a more contrasted format than the yellow

by absorbing some information that appears witih the other filters. This

facilitates the interpretation by emphasizing a smaller quantity of data.

It is important to remember that the other filters must be used to obtain

the information screened out by the red filter,.

-•..-•• •-• Another parameter which has to be considered is the film

sensitivity to the different wavelength radiations. The aerographic in-

frared film with Wratten 98B filter is sensitive to wavelength longer

than 0.7u. This band gives the plant condition from the physiological point

of view. Thus for the healthy dense plants the response will be dark on the

positive transparency. If the plant is not demse enough, therefore,

permiting the film to receive soil high reflectance in the visible portion,

It will be presented as light red. If the plant looses infrared reflectance

the positive image will be more cyan to green. Figure IV.32 shows this

reaction on the area covered by dry (Melinis) which appears green on the

color infrared inags and dark on the black and white.Areas co/ered by(Inperata)*

appear pink to white because of the lighly dense canopy cover on the colored

infrared and, in light gray scales on the black and white infrared image.

The high moisture content areas appear in darker tones because of the low

soil reflectance. A comparison between the plant reflectance patterns, and

- 231 -

filter transmissions curves allowed the establishment of an interpretation

key to obtain the requested results.

The color infrared film has three sensitive layers. The

bottom one is sensitive to red reflected radiation and responds in magenta\

color on the positive transparency. The middle layer is green sensitive

and shows up yellow on the positive image. The top layer is infrared sensitive

and shows in cyan on the positive image (6).

IV.8.4 - Results

Figure IV.33 shows the final map of range!and capability

for grazing. The area could be classified into four classes:

Class A - Areas which are able to receive animals for grazing immediately

during the rainy season. These areas are covered by such palatable

species as Capim gordura (Melinis minurtiflora), Coloniao (Panicurn

maximum), or Napir elefante (Pannisetuan purpureum).

Class B - Areas which are not available for grazing because they are occupied

by unpalatable species. These invaders are mainly Sape (Imperata.

brasiliensis), Vassourinha (Bacchoris rufescences). Batatais^ ™ "™ ™ ^ ^ * " ™ ^1™ *

(PaspaTurn notatum) or other bushes. The range manager can think

about improving his range by control ing the invaders.

Class C - Areas which are in need of land reclamation before grazing because

they are suffering from drainage problems. These are classified by

- 232 -

Panchromatic Filmwith

Yellow Filter

IR-Aerographic Filmwith

Wratten 89B

Panchromatic Filmwith

Red Filter

Aerochrome IR Filmwith

Orange Filter

Fig. IV.32 - Multispectral photos of the same area in four film/filter

combinations.

233 -

u

coCO

COCO

oz

oUl

COOT

o

COto

a.<3E>-K

' JJm2<uo

UJO

itoK)i

o

- 234 -

vegetation type cover. Taboa (Thypha sp.) is the indicator of

high water table soils.

Class D - Uncovered areas which are occupied by roads, water channels or

buildings.

On the map there are some areas which are presented in a

mixture of Class A and Class B, this means exactly that they are covered

by a mixture of palatable and unpalatable species. This highlights an

important decision to be faced by the rangeland manager: whether to control

the invaders first and then institute system, or to use the area for grazing

directly, considering that it has 50% palatable plants, or 50% of the carrying

capacity of an area of Class A type.

Two unclassified areas are shown on the map because of the

appearance of cloud shade on the photos.

Table IV.4 shows the detailed classification into 21 categories

and how they are summarized into four classes.

IV.8.5 - Conclusions

x

Different plant species could be identified by multispectral

remote sensing during an adequate period of the year. The combination of

information obtained along different wavelength bands provides essential

data for the range manager to evaluate his range and to establish a suitable

policy. A red filter with aerographic plus-X film (Panchromatic) gives more

contrasted information, in this period. Plus-X film with yellow filters

- 235 -

provides a general idea about the green cover. Infrared black and white

and color infrared films give specific information about plant condition,

stage and maturity and show if the reflected surface is a living fresh

body due to the high reflectance of the cell, walls and shape, in this

band. Multispectral analysis gives integrated data. Choosing the right

season for multispectral remote sensing operation is the most important

factor to get useful results. This choice depends on the range manager's

requirement for grassland evaluation. Application of remote sensing in

range management decreases the number of specialists needed, decreases

the time necessary for data collection and analysis, and replaces a major

part of the field work to be done, .but requires highly efficient inter-

pretation, of the images. Remote Sensing techniques provide the range

manager with real time information, but he should become acquainted with

how to use this data to optimize his rangeland use.

IV.8.6 - Further Applications

More research and publications are necessary about the

application of remote sensing in the range management field. Spectral

propertiesôf plants in the range are required to enable the selection

of the right sensors, altitude and combination between film and filters for

each purpose are required studies. Estimating the carrying capacity,

.grazing,pressure and calculating the pasture productivity through the

application of remote sensing are important further applied research.

- 236 -

O

onQ

a.•f.o0

UlCD

Of

u_o

(O (A

-o u cI/) V-i- l/> -O

0) R) ••-(O tO O 3

CO 3 Q£ CO

CO CTi O r—r- i— CJ CM

O) t. t— T3jz <v v> v -o•»j > x in at'r— tO C -I— -I— X tf>3. o o EC •«- a>

^~ V) "r— E ^^(O <O *O T— <O r— t/> E:

i_ j_ •»-> v. jc t. cue <aa. Q--CJ a. u O.-M co •— 4-> m

•—it—i i. >—i ca i—i 3 co 2; 2 ex.

10 r oo o> o - CM oo

X C-r- <Ui- i— >

•o •— , Ci— -r- >,Cvj c:ai a> o t- o o «o2: z: oo> ooa.

^ CM CO - in

- 237 -

IV.8.7 - References

1. ARNOLD, R. D., "The Environment - Where Do We Stand? From the Vantage

Point of Space". J.Range Management 25:413-414 (1972).

2. BLYDENSTEIN, J., "Developing Range Management in Latin America", »h

Range Management 25:7-9 (1972).

3. CARNEGGIES, D.M. and J.N. REPPERT. "Large Scale 70mm Aerial Color

Photography". Photogrammetric Engineering. Vol. 35 N9 3:249-

257.(1969).

4. DRISCOLL, R.S., et.al. "Identification and Measurement of Herbland and

Shrubland Vegetation from Large Aerial Color Photographs"

Proceedings of the XI International Grassland Congress:

95-98 (1970).

5. FREYTAG, H. "The Hasselblad Way". The Focal Press. London and New York.

423 p. (1973).

, - 238 -

I

I

I

I

CHAPTER V

GEOGRAPHY

- 239 -

CHAPTER V

GEOGRAPHY

V.I - ACTIVITIES OF THE GEOGRAPHY GROUP RELATED TO THE

USE OF ERTS IMAGES

The Geography Group using the ERTS-1 images for natural

resources studies developed its activities in two phases. Firstly, the

black and white images, of the four channels, were analyzed and the group

concluded that for the type of Geographic investigations carried by the

group, the largest amount of information was contained in channels 5 and

7. This conclusion was based on direct visual inspections of the earth's

surface spectral responses on ERTS-1 images in the scale of 1:500,000.

Thus, the interpretation of the images followed the conventional procedures

used for regular aerial photographies and the results were checked against

available cartographic and bibliographic information.

At a second phase it was possible for the Geography Group

to develop two projects using only the ERTS-1 images. Their preliminary

results will be presented next:

*

V.2 - 6EOMORPHOLOGICAL MAPPING OF THE UPPER SKO FRANCISCO RIVER BASIN

V.2.1 - Introduction ..

The area studied is covered by the Belo Horizonte Chart of

- 240 -

Fig. V.I - Localization of the Belo Horizonte Chart in the Brazil map.

- 241 -

the 1:1,000,000 cartographic coverage of the Brazilian territory (Fig. V.I),

This.specific zone was chosen because it allow the concentration of studies

of geology, geomorphology, vegetation and soils presently being developed

by the SERE Project of INPE. It displays an interesting diversity of

ecological characteristics and consequently can be considered particularly

suitable to generate identification and interpretation keys for geomorphol-

ogical and natural resources surveys based on remote sensing imagery.

V.2.2 - Materials and Methods

.^==- ... :J, r=,;.~.: rThe basic material used to prepare geomorphological maps

were the ERTS-1 images on the scale 1:1,000,000 in channels 5 and 7. A

mosaic was also done for the area under study lin the same scale and

respective channels aiming to obtain a global vision and the articulation

among the geomorphological units to be mapped.

:As auxiliary material to support images interpretation use

was made of geological, topographical, vegetation and soil maps as well as

available bibliography.

Channels 5 and 7 were considered the most useful to geo-

morphological mapping because channels 4 and 6 do not present a contrast

level to make possible the geomorphological uirits distinction.

It was observed that channel 7 furnishes information for

mapping the different types of scarps, lineaments and other structural

- 242 -

sir-i **«—=» —

features. The water bodies appear well delineated in this channel

facilitating their del imitation-by water absorption in the infrared" "

band of the electromagnetic spectrum.

'..,.. . Secondary drainage is mapped in channel 5 through the ciliary

forest that absorbing radiation in wavelengths from 0,6 to 0,7y will appear

dark.

Comparing these two channels it is possible to have an idea

of the vegetation cover. The grey tones at the {Different 5 and 7 channels

will indicate the grade of vegetal density. Through textural differences it

is possible to obtain information related to relief erosion degree.

Drainage, grey tone and texture jpermit initial delimitations

of geomorphological units.

For each type of desired information an overlay was done.

Thus for each image we studied the drainage system, grey tones, texture

and structural characteristics. Superimposing these overlays it was possible

to delimitate the geomorphological units.

N.

The additional necessary information of altimetry were

obtained through schematical profiles in the Images, compared with topo-

graphical profiles of the Belo Horizonte Chart, Through the comparison of6

data and delimitated units in the images it was possible to obtain the

altimetric amplitude of the relief area.

- 243 -

V.2.3 - Interpretation Criteria

The adopted criteria for geomorplhological interpretation

were directly based on image characteristics and bibliographic examination.

a - Drainage System

. Jfv,!=,., - - - . : _ Through spatial drainage arrangements it is possible to

infer a series of important facts to relief forms mapping. A drainage

system of a region reflects its structure, permeability, topography,

vegetation, etc. .-..,..

The ERTS-1 image permits a synoptical vision of the

drainage system thus furnishing a preliminary separation of geomorphological

units. A dendritic pattern, for instance, provides information about a

region's declivity, permeability and predominant fluvial geomorphical

processes.

b - Structural Characteristics

To the delimitation of geonrorphological units the

presence of faults, scarpments, circular structural surfaces, folds,

alluvial slopes are important when associated to a drainage system.

c - Information on Morphoclimatlc Systems

The interaction among the raorphoclimatic systems and

- 244 -

Structural conditions and lithology define morphogenetic systems. So in

,,: .macrocorapar.tments, from local variations of structure and climate, it-is -

possible to infer morphogenetic systems.

d - Altimetry

The relative compartment's altitude as well as the

uneveness amplitude are important information to a final geomorphological

interpretation composition. These data could not, at this moment, be

quantitatively defined in the image, but could be infered by the behavior

of other information which the image can offer.

From these criteria an interpretation was done of the

mapped.units and a compatible geomorphological legend adopted with a chosen

scale. The mapped units are described in terms of the characteristics

presented in the image and of the possible corresponding characteristics

1n the soil.

V.2.4 - Results

The geomorphological mapping using ERTS images, considering

the scale and the absence of stereoscopy, restrict itself, at the moment,

... to. the..great relief-structural compartments or macrogeomorphological units,,

permitting the .principal types of inference of the dominant morphogenetic

processes.

- 245 -

As an example, to have the interpretation of two images in

_ ,.the channels 5.,and 7,of .the. area under study,, figures V.2, V.3,V.4,V.S -

were selected by their great potentiality showing practically almost all

the geomorphological units types occuring in the area comprised by the

^.- Belo Horizonte Chart. The area corresponding to these images is shown in

figures V.6 and V.7.

... ..:,=.,.:; .-. -.,:-.-:-= .-Generally speaking the area covered by the Belo Horizonte

Chart presents a satisfactorily differentiated relief shaped in Pre-

Cambrian and Paleomesozoic rocks.

Through image analysis it was possible to find large

compartments as well as to differentiate the regional sub-units (Fig. V.8)

listed as follows:

A - The Plateaux and the Espinhago Range (Serra do Espinhaco)

This is a well defined unit in the image because of its

distinct photomorphological characteristics from the rest of the area.

Due to its structural and geomorphological complexity there

appears several texture types, tonalities and tfie drainage pattern varies

from sub-dendritic to parallel.

Geologically speaking it is forraed by soils of the Minas and

Lavras Series constituted by til its, quartzites, phyllites, sandstones and

metasandstones. -

- 246 -

_ uaa5-30i89SEP72 C SI7-2I/U045-I5 N SI7-23m0<J5-88 HSS 5 R SUN RZ06I I88-0666-R-I-N-D-2L NRSR ERTS E- IOT8-I23H-5 02

Fig. V.2 - ERTS image E-1048-12.314 - Channel 5.

OBIGINAL PAGE ISOF POOR QUALOT

- 247 -

UBa5-30l09SEP72 C SI7-2I/W845-I5 N SI7-23xU0^5-08 nSS 7 R SUN EL<>7 RZ8S1

Fig. V.3 - ERTS image E-1048-12.314 - channel 7.

ORIGINALep peea

- 248 -

- - 0l U8a3-33lISHUG73 C SI7-l2/Ue3-5- N S17-1<VU8"3-"8 nSS 7 R SUN EL"I R2053 188-5^21-R-1-N-D-IL DflSR ERTS £-1389-12255-7 01

Fig. V.4

ORIGINAL PAGE ISOF POOR QUALITY

- ERTS image E-1389-12.255 - Channel 5.

- 249 -

CoIQ QQI

I6RUG73 C SI7-I2/U023-5«"N S\1-\?'U&>3-a8 tlSS 5 R SUN EL/SY RZ853 I88-5"2I -R- t-N-D-2L NflSfl ERTS' E-1389-12255-5 01

- Fig. V.5 - ERTS image E-1389-12.255 - Channel 7.

POQR

- 250 -

Westerly it is delineated by a distinct fault scarpment

forming an abrupt contact with the BambuT Series rocks. This unit represents

the water divisor between the rivers flowing to the Sao Francisco River and

the ones draining to the Atlantic Ocean, represented here by the JacuT

River drainage and its tributaries. Easterly the drainage is represented by

the Jequitinhonha River and its tributaries. Although it appears as a macrp_

unit, it does not present a lithologic geomorphological or paleogeographic

unit presenting individualized sectors as:

1 - Relief with Parallel Crests

The relief appears in the image with rough texture

showing great variety of alignments, with predominant North-South

direction. Their grey tones are dark in channel 5 and lighter in channel 7,

although appearing in this channel as dark gray when compared with adjacent

areas.

The drainage system is relatively dense and many rivers

present a clear structural control.

This is the sub-unit which takes up great extension in

area and it is a typical Appalachean relief with phases of the Jurassic

type or iriverted, where appears a sequence of parallel and sub-parallel

quartzitic crests that resisting erosion give to the group a montainous

and sloped relief, associated to a fault, fractures and folding system.

III

- 251 -

I

I

I

I

I

I

BII

Fig. V.6 - Localization of the studied area in the Belo Horizonte Chart,

-Co . -:r-/> ' .-.•

3 •• -Y T-- •— •« .- •- .-• v -. —^/-; Af V4* ~ ~ * ? * -: "- -^ '•--• - • • -.^•/:-\^V- • . :-\'--^\s'-Z '

-^->:-^N^. • «~- .-• • •-."•: I-\\v- ..V; V ; ^ \=L^ \ • - . - . - -:, -S Sû.7

^C."".".7 -. _. . y«5^V<C.:_"- • -•'^r^f^'r^-,'^^''<^'S^ • • • ' . r-^---v -4- '^*r-JvV--—••;--:•<!!^^f'^^-*^-.v'-i.->v-r^'"v.-—*v A• *&?; \:i:" ~, >*>^r'~ ^k: '^z-^j^îr^r/r^-. /.::•-i':.--/--"^ •- . . -^ * ' ? - , - • - - - •

1 1 - :': ; • _ • • - - ;.—;•;' :"^" "*" -.'.- \- *-. l^/'l*-"^.." ~ —

i ^Ti^j^:',.- -,•"• "r-v ;-,~'.'l s ':>\'r"\(iv*; ',' ,-v •:••'•'. : '~~^ •

W-

i» ; -*,>_-"t:\_ * • • • - f"

_N •- -- "" •

RiO D£ IANERO

&-,=:-

-^ ^_ * .. c; i '•Sjg^LV'r" ' r- g "^T .'-^"-=r:->7""~"^7"~"" ™ ESC«> -J"<

I:IO»**1

asssr-'

—1

v - • .r ~ -*-•. "-••: S-"— i"" -»V 'W - » -1 tV^ •-. •- : •--'; • - • : ••' - ..«-i-- ^r^ •'•rsr'- V'-^-"~ :--

' i -•-. *' ~>- ^ ~ ':% -^

• . T^<-. -. - ' •-iv ' «'

-•y. C-r . S^C^^"^^-rT;---^^7"^^f/r?;- '^ ,/" .---' "-. . .'-^*-*1^.

~ . J - f • ~ *~$ \ ' ~~ .*" «*r'-» ^v XT™" " " "" ^ »*- - M1"* ;^ * •'•^•'•U . v^~-^>.j " " \--rm i - *** ^^ vj * ' ._ *... •"—•

/

^^r~-A.:^--;^ i

xrt-•OCV-*»*-««0

- 252 -

13°OOW

I7°OOS "

• Vv ' • • • >Y' / K AC?/ r

Wx te^rÂ^Jp ,[<• r1^^ ^^f <<cPJ ;e«i»»--X- '= • i.w'X*4-. N}-.—v?-v

44°OOW 43*»OOW

- - —' ^ N'- I s"^ *io.' >

7>;-^JW\tfsS-'"J? "~~"V \^^î£T^^~~^ -^ '-••• •*'^Hô" y02vMa.i:.. ' . ; M :

»,•!.|O, " -." "/ ~- —» *'— * .^ _^ - /* _— ^•..-^9obn«'iro<Srf ~\

ftf*1' ;A- ix** v-4^~?*" •:-- ^^Nur-J-'

JjJ.N^BlTf 00 S\»^*^?\y-\ ***°*f£;*i/\)r

• '. t^ frm«.

44?OOW

4 and n9 E-1389-12255.

- 253 -

t

Fig. V.8 - Geomorphological outline of the Upper Sao Francisco River Basin Region.

•

•

GEOMORPHOLOGICALOUTLINE OF THE UPPER

SAO FRANCISCO RIVERBASIN REGION

4^>j

i *ttd

\ i i • • e. ; 13 ^wI i I ^••\\ 1 • !• • .r -\ 1 M.W ; ill? ^' ' ; > »3 ^^HÎ •> : /• .< oKA i

4STOO W

.s-3 • £>

iT'bo's

I7?30'S

M i !Hi!

LE

LE

43.30W N

RELIEF WITH PARALLEi

THE PLATEAUS AND Th

INTERMOUNTAINOUS DEP!

SEDIMENTARY PLATEAU

TABLELANDS

INTRUDED VALLEYS IN

SYNCLINAL

ANTICLINAL

EROSIVE SCARPS

FAULT SCARPS

RESIDUAL PEAKS

DRAINAGEc

WATER DIVISOR

*7 JTQ

L E G E N D

RELIEF WITH PARALLEL CRESTS ' - ' . ' . • . " ' . " .. . \

THE PLATEAUS AND THE"SERRA DO ESPINHACO'(ESPINHACO RANGE)• ' j

INTERMOUNTAINOUS DEPRESSIONS : : .v f

SEDtMENTARY PLATEAUS

TABLELANDSLEVEL I

LEVELII

INTRUDED VALLEYS IN THE TABLELANDS

SYNCLINAL

ANTICLINAL

EROSIVE SCARPS

FAULT SCARPS

RESIDUAL PEAKS

DRAINAGE

WATER DIVISOR

- 254 -

The "Serra do Cabral", although spatially separated from

the great Espinhaco structure, is part of it through similarities of

lithology, structure, geomorphology, drainage system and image photogeo-

morphological characteristics. It is formed by two big anticlines and one

syncline, besides other foldings perceptible in the image. In the axis

of these two anticlines appear quartzits highly resistent from the Lavras

Series, whose altitude is reflected in the saw format.

2 - Intermountainous Depressions

They appear in the image with rough texture and light

grey tones for both channels, although lighter In channel 7. Distinguish

itself perfectly as an eroded area, emerging older subjacent rocks which

are delimitated by clear erosive scarpments ire the contact among soils of

the Minas Series, formed by sandstones, clays, 'conglomerate and pebble of

metasandstone.

A clear example seen in this image is the Itacambirugu

River and its tributaries depression, whose suft>-dendritic drainage is

truely representative.>

These ssarpments were origimated from erosive processes

and constitute themselves in a water divisiorr Ibetween the Itacambirugu and

Jequitinhonha River Basins.

- 255 -

B - Sedimentary Plateaux

This unit was identified in the image through a very

dark gray tone in channel 5 and light in channel 7 with smooth and uniform

texture and with low drainage density and 1 imitated by well defined

scarpments.

Geologically it corresponds to sandstones dated from

the Cretaceous Period, highly porous, horizontally disposed furnishing

the necessary conditions to the appearance of structural tableforms

surfaces.

Spatially present itself descontinuous and is an

example for more intense surfaces westerly localized in the Sao Francisco

River Basin.•

1 It is covered by denser "cerrado" vegetation than the

adjacent areas, infered through the light tonality in channel 7,representing

the high vegetation reflectance in the near infrared band.c

v.

Topographically represents surfaces maintaining

altimetric levels from 800 to 900 m.

- 256 -

C - Tablelands

They were differentiated in the image through a light

tone in channel 7 and a dark one in channel 5. the rough texture on this

unit boards clearly shows the erosive scarpments that constitute the link

element with inferior levels of the tablelands.

Over all this unit the drainage presents a sub-dendritic^

pattern suggesting the presence of horizontal or sub-horizontal rocks with

uniform resistance and with a smooth regional slope.

Geologically speaking this is a unit formed by rocks of

the BambuT Group, where limestones, slate and conglomerates emerge.

Morphologically it is represented by plateaux strongly

eroded with occurrence of erosive scarpments and forms of Karstique relief

in the areas of calcareous rocks outcrops.

Among the Espinhaô Range and the cretacic sedimentary

tablelands this unit presents two distinct altimetric levels: one westerly

located in the "Serra do Espinhaco" and at the east of the sedimentary

plateaux and the other intruded between both, at an lower altimetric

level. This compartment division is specially done by the presence of

scarpments at both size of the Guavinipa River Valley.

An evident image aspect is the tablelands at west of

- 257 -

the "Serra do Espinhago" which is less eroded than the one located at east

of the sedimentary plateaux.

D - Intruded Valleys in the Tablelands

i

This unit was delimitated through its smooth and uniform

texture, through the localization close to the principal channels and

through low density of secondary drainage.

Geologically they correspond to rocks of Vazante Formation

which are deposits of the Quaternary Period, already consolidated and in

an reimprovement phase. They follow the Sao Francisco River Valley and

the valley of their main tributaries.

The horizontally layers form smoothly inclinated

tableland reliefs in direction to river beds which maintain themselves

at middle altitudes of 500 m.

Close to the Velhas and Sao Francisco Rivers in thec

limit of this unit through relatively abrupt scarpments.*x

*

At the right border of the Paracatu River also appear

small scarpments while at the left border they do not appear.

- 258 -

V.2.5 - CONCLUSIONS

From the preliminary results of this geomorphological

mapping one can conclude that:

- The MSS/ERTS-1 images could be considered as a valuable

Instrument for a synoptical vision of the relief

compartmentation and of the structural influences.

- Related to localization, extensions and limits of the

geomorphological units the images permit the introduction

of modification in the already existing maps.

- The images offer possibilities to infer about morpho-

climatical processes and their dynamism through repetitive

analysis of the same area.

V.3 - HYDROGRAPHIC MAP USING ERTS IMAGES

V.3.1 - Introduction

The hydrographic system is subordinated to geomorphological,

structural, Hthologic and climatic conditions of a given region.Therefore*

it is a valuable element for environmental research. A detailed study of

the drainage system furnish a series of important information to the

determination of a region ecological scene. It is considerably important

in the preparation of hydrographic maps as an auxiliar element in the

investigation for other parameters.

- 259' -

This map was initially made aiming lo support geomorpKolcaical

interpretation of the area covered by the Belo Horizonte Chart and

exemplifying the drainage mapping. The images used corresponded to the

same area used to exemplify the geomorphological mapping of the previous

work (Fig. V.7). <.'.»

^ 1 'Through the above mentioned map it v;as possible to have an

idea of the area compartment division in function of drainage and of the

drainage patterns, •

fill 'the present moment only a qualitative analysis has been

done of the hydrographic system, although quantitative..studies will be dojja

with the use of Hortton and Sthraler morphometric indexes.

V.3.2 - Drainage: General Organization

Through visual analysis of a drainage system (Fig. V.9) it

was possible to distinguish two great hydrographic sets separated one

from the other by the Espinhago range and plateaux.* * c

V

The Espinha£p alignment operates as water divisor separating

the drainage oriented to the east (Jequitinhonha .River and its tributaries)

from the westerly oriented system (Sao Francisco fciver and its tributaries)

The rivers flowing to the Sao Franiisco River present a

- 260 -

rr bds

7°30'S

I

I

I

I

I

Fig. V.9 - Hydrographic Map of the Upper Courses of the Sao Francisco andJequitinhonha Rivers.

IIIIIIIIIIIII

i Ac's

HYDROGRAPHIC MAP OF THEUPPER COURSES OF THE

SA'O FRANCISCO ANDJEQUITINHONHA RIVERS

o '45.30W

e '45.OO*

46°00'W :oow

10 O n to so 40 to

4^30 W

e'bo's

. — — I WATER DIVIS

- 261 -

superficial dendritic pattern and a high density of secondary rivers. The

Jequitinhonha River Basin drainage already present itself less dense,

reflecting even in the less important tributaries a great structural

control with a sub-retangular pattern.

Correcting the drainage map with the geologic map of the

area, ft is possible to verify whether the drainage is denser and dendritic

in the areas of rocks occurrence of the BambuT Group. In the areas of

calcareous occurrence Karst topography can be found. The superficial

circulation in the Karstique region equally elaborate deep cannyons as

the one formed by the Velhas River in the region of Santa Lagoon.

Where the climatic conditions do not permit the appearance

of Karstique relief, the calcareous works as a non-permeable rock making

possible a rich superficial drainage.

In -the recent alluvium areas from the Vazante Formation,

the drainage is less dense, perhaps reflecting an increase in the porosity

of the rocks. 'e

In the place where occurs more ancient terrains of the

Brazilian Shield the drainage appears less dense in the images. This,

however, does not totally reflect the reality. What happens is that in

these regions it is difficult to obtain drainage details through the

images which allows to map only the most important rivers. This is due

to two facts: Firstly this region present itself covered by humid forest

- 262 -

diminishing the contrast in the targets and this reduces the image

resolution. Secondly, this is an area where the drainage branching is

great, presenting small and narrow rivers, more difficult to be detected.

\

V.3.3 - Conclusions '

I

By the drainage analysis using ERTS-1 images it is possible

to conclude:

- The image provide information very close to reality,

attaining a level of detail much better than the ones

of the chart in the same scale.

- By the facility of drainage mapping, the ERTS-1 images

are favorable to hydrical potential studies, as well as

to the study of fluvial channels with the objective to

plan river transport.

V.4 - DEMOGRAPHIC INFERENCE USING ERTS IMAGES

e

V.4.1 - Introduction* V

Among the different ERTS satellite image application

possibilities, those concerned with demographic studies seem to be some

of the most appealling.

- 263 -

In general it is possible to observe that urbanized areas

in ERTS images are well defined which really facilitates the delimitation

of these areas with fair accuracy. This fact inspired us to verify the

viability of the ERTS images utilization in demographic inferences in the

Brazilian territory.

I

At present, Brazil has approximately 100 million

inhabitants, with a population growth of 2.9% per year, in the period of

1960-1970. Demographic Census are done here every 10 years, hence some

appropriate techniques are necessary to estimate the population in the

inter-census periods as well as to accompany the main urban expansion

trends. Frequently it is noted that the Brazilian population is

concentrated in highly populated agglomerations. The percentage of people

living in clusters of more than 10,000 inhabitants has been increasing

considerably in the last few years.

Due to the intense urbanization process presently occurring

in Brazil it is necessary to keep track of the increasing growth rate

within the urban areas as well as to monitor the expansion of thec

urbanized areas.

*

The frequent ERTS satellite data acquisition makes possible

the detection of the urban expansion tendency and the observation of

regional developments, making it possible to update regional planning

related to urbanization studies.

- 264 -i

Trying for the first time in Brazil to use ERTS images

applied to population inference, this work verified the correlation

existing between the data of urbanized areas, as infered from the images,

and the respective population data, obtained in the last Brazilian

Demographic Census (1970).

Images used in this work were selected according to their

clarity in defining urban areas.

V.4.2 - Methodology

The data used in the development of this research were

obtained by ERTS-1 Satellite MSS Sensor. Channel 5 (0.6 - 0.7y) was used

because it was the one which best identified unban areas. The urban areas

present high reflectance in this channel companed with their surroundings.

Through the radiation density cities could be delineated with a relative

accuracy. An example of two utilized images in the work are shown in

figures V.10 and V.ll.r-

The scale of the images used in the determination of the

urbanized areas was 1:500,000.e

Cities were identified using 1:^,000,000 map of Brazil

published by the I.B.G. (Brazilian Geography Emstitute).

a

The determination of city areas in the ERTS images was

made with a planimeter. The arithmetic average of 5 area measurement

was used.

I- 265 -

..• \ '•s-';r:V.7s-'':=' "'~'-'*'-r:" *'-•• *'; ''s ••-•/•'..- •': * :" • .: cSS&S /•*?ir :-: \'•.i-. t -" • -: .£>-3 S";- x* ? ;* ; :?•^ , •--••.-.-^Ci?:1- -v .•'-.- :iv *-.» ;. .-...-'/•:.-^ ;

I'J0a7-30l UB«7-88I U016-30I

38JUL73 C S22-52/U8'»6-'>9 N S22-53/U0a6-al OSS 5 R SUN EL33 flZ8«6 188-518^-3-1-N-D-2L N^gR ERTS £-1372-12333-5 02

Fig. V.10 - ERTS image E-1372-12.333 - Channel 5.

—I—

I!i'III1iii

f*J»«- /i

V«,

' "*»-w'-'»»'

'%-.

"''\-

*"

»>.,

% - . ' - • ••>:• -"v^--.

..~f'..;«v^^--^'^

^ .' ;. •< . .•'.v v •

<*.

*' ^>v, -'i

:^>: '\"-'- . "*. .^z^cJ^^^'^. *±^£r' *

"'%*

v--^'

89SEP72 C S2I-4I/U046-I9 N S2I -<?2/U846-l<> flSS 5 R SUN £La& RZ058 l89-B666-fl-r-ff)-D-2L NflSfl ERTS E-1048-12339-5 02

sI

Fig. V.ll - ERTS image E-1048-12.330 - Channel 5.

- 267 -

The area data were correlated with official statistical

population data of the Brazil 1970 Demographic Census.

As the first step to analyse the data an attempt was made

to establish the correlation index between two variables: the area and

the population. Then a linear regression was used in order to estimate

the population corresponding to a given urban area. The next step was

to apply the Student Test with 90% probability.

Intervals were established from the data obtained by the

simple linear regression methods with a statistical distribution of

population resulting for each area interval.

V.4.3 - Discussion of the Results

This work is still in the first phase of development and

the results to be. discussed here should be considered as preliminary ones.

From the ERTS images selected for analysis, 280 Brazilian

cities were identified (Fig. V.12). A list was made with information about

the place of each city by state, data on spatial area obtained from the

images, and population totals acquired from the last Brazilian 1970

Demographic Census. The 280 Brazilian cities identified vary greatly with2 2area ranges of 0.5 km to 45.20 km and population totals from 559

inhabitants to 185,000 inhabitants.

- 268 -

B R A S I L

INITOTO

PCJANEIRO

Fig.V.12 - Brazil map with the localization of the studied cities.

- 269 -

o>•*h-3 ir c t\ & c\ a c - — f - < " t c r - c c » - K t - f r ' f C C — c v c t i r — uro ^ — t~ c — <r tv « c i r » ^ i r ^ i r « i » « ^ « c f - « c .«f<3«''c«r«cr-coci».<c^f*i«.ir'-^c*.ciyK«-c"ce<c«'l'«-ir^D o c f i r « t ; / « » c » T < r c c c - c s r f c - < r e » - « < ' ' C * - < f O B « r « e » « - w ^ c c « \ . « - t r c e c o c v i r i r — e ' c i « ' « » - < a t vQ. .- .- IT c- IT — e\ c. iv. r <•• « «\ K i* r — v < c « c i r ( v c ' f v i r r - p « ^ r k . W ' - e N . F - r w K . t r i \ . u - c c >»ivO «^ — ix < o <\ c a <r . o a (\ tv ir K — t>. a — — oQ.

e\7*

5c c c c c c c c c e c c c c c c c - c c r . c c - c c c c c c e - e - c c - c - c c c c c c c - c c c c c - e c c c

2 .\".f.". *. *.c."".'.c.. . °.°.*.c..*.e. '..c.^c.^°."". e..cv..e..x.e.* c.".fv..ir..ir.evcVr..'. <c.c.' c. * c..~.*l.<£..".ty ~- <T f v c x o - i r - w e i r i r c — « « < v < ^ < T K < r / " « ^ c ^ - f v < r « t ^ f \ . t r e r L * " c x * * " ^ « * i - ^ « c * r * r < \ . < c * - c \ ( r<: - - « - . - , - . « » . ~ - -

OCOCM Ul c

c

». •- « « r u-z »- -i e •— o< z o ec _i

i« e»i O

« «r c « ui— </; i/> a c *- '

to O «z o- >-

a: a o C «-

u.- « c c.

C «?I o c-u </>

r u. a. -

u- u; =• c -.

- 270 -

Q.

g

ire

«f ir er cir c — c.

— « i* c K — e f cr cc r i r e < K K .a c c v f ^ « v < c < c««-rv

c e c r e o c - c c o c - c c - c c e e e c e c o c e c - e c c - e c c - c - c c - c c e - c c e c c c c c c e e e« c c « t « c o < x o - o t \ . « — CfM<*»

«*LU

<c— — «v

P f «•• ir « c': c tr «. «r c c•rtwie-, r - x t o a o r o . wv;ct3: ec T i </>

I 5

- 271 -

r\Q.

*•—CM

c . c c c c c c o c c e c e o c o e e c c c e c - c - c c c c o c e - c c c c c - c - c e c c c c c c c - e c e c -

^ x i r « n « / , o i r « / i « r m < T 3 TO ?• T to IT r T r i « / ! « T « / ) « r s ' < / % « / > T : « £ « / > v 5 a . « / > i r x « / i < / 5 i / > T i « T t r r «; </; a. v. «/»t—

v>Ul

« c.

« «r. ir j« _j ui vt " <— ~

^f I X *f — < ^ ^^^ ' O I*/ U^ ft. 2. A. (^ O

- 272 -

o-o0.

. - c v n c f v - f f t v i r c c i - . — t r p f f c v c r O i r — C K « « e c ' * . N . « ~ * . i v i r i r H . — — « » > « e — C ' f ' - ^ o t c K - i r e * ' -rfrcvcrir«r—'fcvKoirc\r i>.f««,-f»-««£c^—o<c—«' r i rc\c»'-«r«N»>.ir — » - e - . - i . - i r a f i r o i . - c . -c — K K c < o — < £ a c a C c r o - c c v « - o « » - ' ' ' * - « ' " < v i r e — i r — »• o or — c « — r- o ir <r cv — ir s tf & c rK o tr — < »r — i~ « cv — — » - < v o — c r « — « r < c ^ C k C r — xc* ir IT <"•. «•>.»» c-. c •-•-«*•. o- o — «• — .-a

CVJc c c - c - c c c c c c c c c c c - c c e c c c c c c c c c o c c c c r c c c e c e c c e c c c c c c c c c*C* if. C f 9 O <C C^ C CV O Q IT* ff ^ tf «f> C* O *C ** «(' ' 9 X C O C? f Q C~ V^ CC >C C^ P^ C* tf C i^- 9 f O P^ l^' O GC N>

(^ tf P* Pv P* P* IV *C If G •*"•' - C\; »•- C ^ ^" *4 CV CV' If" »* If* f*! <*^ ^" CV CT CC ' CV f^ ^^ tf 1 . Ct CV *C f'' »C CC »•• !*"« ^ *f f . ^* CV ^™CV»- CV ». ». ^ ^. •- «.

t r c a c o - c a - c o f c t - « > c : «r tr c. c cr.c c e « o r o t r e ex s i / : i / > a . < ; r x z 3 i / ) S T i r

ac. •T W

~ >- « c> -a

>- i/iui •-z _<

cin vp.c x c:

I/Iu>

7«(C

U'

IA.<

C

UlO

oc

X.«z

o c z

«z*«e 02 a. «o s zc < / > « r » z » i o. a. 0z «/•. c u. a c: •«*. «r. — ~ ~ « « « r i/;ci/: z

*- U - U J ^ . ? S < ? ~*O ~ C ; i - ' T ~ O U . «!*•*— C Z

* - i ^ ~ ' C ' ; / > > s * - u . O T ~ — « ic^ — — > r < u ; _i/s^« cc: 3Ta ' c z u-u. cu . -u . c r c - ' t r o ^ * - ' — < C u . r r a * - — *u. r a — —

_ * ? c ^ « - « r « < « - 4 - s « < ; c - t r . « / . i ^ - i ^ Z C< j« -*3o_ ' a . ac : zc . i : z o r m i / i c c ^ T ^ z2 = = c n . o . a . a a c . = . a , a . a . c a a c ; a . ~ a a .

SBRSS!

- 273 -

cv *• <•• c <rf ' - r ^ c v— « • « ? « : <e t a —

CV ~ •• ^ •» CV

c e e e c e c c e e c e e c c e c e c c e e c e c c c e e c e e - c c e - e c c e c c c c e c e e e c e

c . o c r c

«• f C

Ul

aucoa

v. _Ju «s a

V.- C VC a a 0 C —a 3f t> s c =

« u. (.• c; a. v.'.ia s f. < «• u «z z or a e a •-c c c a c c a

« a. u

•2•- e c c.

c-c cO G

u;«/•- CC •-C I/•: u.2 a;a. u

«r n-«t c1 eo c c i~

c C- c c ««: c e- — o- •- «•«/!U'«r«?<«rtfr —v'^strtrC—'OZ'U.

«/>«^

:' C«r

e- o e _« «. u.= ir K, •/: ac c rv» r-

« tru-r ue c

~TZ o«/:~«/r— <w _• Q. u « - < = « — —« « v i t e r ur

_ c o . « r r s c««r« r e <-«>e<.- _/ —

xir <c ec — «_-a «a.c : c c c c cc-^c-rc.- c- w«/- <c

« • v c r « r _ i c - c r « r c c r c e>-«i_ cr~u.^c — ccccc.a f c - s c » « . " - « s r u c c u« • — U . ^ ~ T S C T « C ' C U I » U ^ C ! t ~

r u u i * _ » - > - H - > -£ « ; i < r c c c — z - ^ r recu. — — •-•-•-•- = «< «a D a a f f A a o c i r t r i r

u i r c * » . * ~ ~ ~ ~ ~ : ^ Q C 2= c - c c c - r : c c c c c c c

ORIGINAIJ PAGE E{OH

- 274 -

c v c - c i r — • f r v c af <. •- c <•• o

c v J u - t — c f v < r c - < u ir K c- op.-«-.(\«- C-- — c r «•

csi

c c c c c c c c c c c c c c c o c c c c e - c c c . c c c c c c - f l c• « » • * • * « » • * *. •» •» • • • » • » • * •. *l «. w •» •» *t •• *• •» «» «» •« *k «• *v •* Mb «

na v c" c k- r c c c c c c ir •- o c « t.* i- c r. c c c c «.- c c *• c ~ <r

v><a

C •- Cc 2

'

U ifc »^ fr - « . «•1T — - ~c c- « «r •

CC V

- 275 -

With the application of the correlation coefficient to the

variables, areas and population, the obtained index was 0.80 which

indicates a relatively high correlation between the two variables.

The Dispersion Diagram (Fig. V.13), shows a linear dependence

between area and population. The dispersion from linearity observed for

some points in this figure probably indicates a peculiar space expansion

of these cities with respect to its population distribution. In these

cases some parameters should be taken into account, such as: regional

localization, topographic characteristics, functional and economic

characteristics of the city, etc.

Later on, with the application of the simple linear

regression statistical method it was possible to classify the area and the

population in intervals, resulting for each area interval a statistical

population distribution. About 21 classes were obtained as shown in Table V.I.

These classes were established in order to infer the

population of a given sample area.

The results obtained for the less populated cities are

quite different from the more populated ones. For the less populated

cities the relationship between their areas and respective population

was found to be quite linear. However this linearity does not exist for theo

most populated areas. Table 1 displays the area intervals used in this

work. The corresponding classification and population data are also shown

II - 276 -

CD

2

215

•5

• «.

ooo8

o o oooo

I 1 i I I

o po 5o oo oS S3

8 ao o

«

3

! I> en

ID

i §! T-

i £• O)• o.' VI•i—; o

ien

, CD

o9

OOo2

oo

- 277 -

in Table 1 and in Figure V.14.

These results (although relatively good for the population

estimate can be considerably improved if more samples are considered.

There is the need of data collection for a larger quantity of cities, and

also the need of other parameters which influence the geographical expansion

of the cities. Parameters such as regional localization, kind of relief,

economic and functional characteristics should be analized and computed

in order to obtain more precise results in demographic inference.

Based on these preliminary results one concludes that the

ERTS images really offer great potential in demographic studies and in

tracking the expansion of urban areas. However in the case of demographic

inference more precise results are necessary and could perhaps be obtained

in a further phase of this research.

The next phase of this research will be developed utilizing

the IMAGE-TOO GE equipment. Through this equipment more precise results

of the urban areas will be obtained. Other parameters interfering in the

urbanization process will be used in this second research phase. The

objective of this investigation will be the reaching of models to permit*

the inference of urban areas population with the best possible precision

grade.

- 278 -

TABLE V.I

f

CLASSES

123456789101112131415161718192021

AREA

0 to 2 km2

2.1 to 4 km2

4.1 to 6 km2

6.1 to 8 km2

8.1 to 10 km2

10.1 to 12 km2

12.1 to 14 km2

14.1 to 16 km2

16.1 to 18 km2

18.1 to 20 km2

20.1 to 22 km2

22.1 to 24 km2

24.1 to 26 km2

26.1 to 28 km2

30.1 to 32 km2

32.1 to 34 km2

34.1 to 36 km2

36.1 to 38 km2

38.1 to 40 km2

40.1 to 42 km2

44.1 to 46 km2

«

POPULATION

0 to 6,838 inhabitants2,941 to 12,7029,222 to 18,65315,720 to 24,71021,801 to 25,32228,365 to 37,13533,654 to 43,46641,775 to 49,84745,228 to 56,26250,958 to 62,70261,787 to 69,15864,628 to 75,62668,034 to 82,10373,818 to 82,88189,552 to 100,92095,486 to 107,74096,899 tp 111,314103,113 to 121,065112,711 to 127,568115,534 to 130,819127,384 to 144,481

CITIESNUMBER

497757291411765514411132111

- 279 -

(AIU

2

oen

o •>>~io

4-> QJ•i- l-i— «

JO<T3 VIjo uio <ol_r—Q. O

O) Cj= a;•u >w) en*->c <o

ri

o" g o

S~ S-<O (U

JO 4->

•t- O*£O) (0

O> Q-.c oI— O.

- 280 -

V.5 - BIBLIOGRAPHY

AB'SABER, AZIZ NACIB, 0 Relevo Brasileiro e seus Problemas, Brasil: A Terra

e o Homem - Vol. I, Cia. Editora Nacional, 1964.

ALMEIDA, FERNANDO P.M., Os Fundamentos Geologicos, Brasil: A Terra e o Ho-

mem - Vol. I, Cia. Editora Nacional, 1964.

COELHO, ARNALDO GUI DO DE SOUZA, Ecologia e Potencial de Estudos com Base nas

Modernas Tecnicas do Sensoriamento Remote, I. G., 1973.

DOMINGUES, ALFREDO JOSE" PORTO - Relevo , Paisagens do Brasil - Fundacao IBGE,

Institute Brasileiro de Geografia, Rio de Janeiro, 1973.

EYRE, L.A., B. ADOLPHUS and M. AMIEL, Photogrammetric Engineering, 5, 460

1970.

HOWARD, A. D., Photogeologic Interpretations of Structure in the Amazon

Basin: A Test Study, Geol. Soc. American Bull.,1965,V.76,

pp.395-406.

HSU-SHIN-YI, Photogrammetric Engineering, £, 449, 1971.

LOWMAN, D. PAWL, Multispectral Scanner Imagery of the Sierra Nevada:

Geologic Analysis, Earth Resources Technology Satellite-1

Symposium Proceedings, September 29,1972, Goddard Space

Flight Center, Greenbelt, Maryland, NASA.

LYON, R.J.P., Exploration Application of Remote Sensing Technology (Image*

forming systems): Mining Cong. Jour 1972,V.58,N9 6,pp 20-26.

MOREIRA, AMEilA ALBA NOGUEIRA, Relevo, Grande Regiao Leste, Volume V.IBGE,

Conselho Nacional de Geografia, 1965.

PENTEADO, M.M. and RANZANI, G., Relatorio de Viagem ao Medio Vale do Sao

Francisco, I.G., 1973.

- 281 -

WAY, J. R., Third Annual Earth Resources Program Review, Vol.1, Geology

and Geography, £, 1, 1970.

YAMANE, T., In Statistics, An Introductory Analysis, A. Harper International

Edition, N. York, 368, 1964.

Ptf/

•*- NATIONAL AERONAUTICS AND SPACE-ADMINISTRATION5 . GODDARD SPACE FLIGHT CENTER

,£-' _ GREENBELT. MARYLAND 20771

" TV'

REPLY TO _ • __„ATTNOF: Code 9 0 2 , • . , , -

I I L t JLO

3 1975

Dr. Fernando de MendoncaDirector General, I.N.P.E.C.P. 515 - Sao Jose dos CamposSao Paulo, Brazil

Dear Dr. de Mendonca:

In accordance .with the Provisions for Participation in the NASA LANDSAT(ERTS) Project, a review was made by the Scientific Monitor of the draftsof the Type III Final Reports which represented the completion of theInstitute for Space Research's (INPE) multidisciplinary program for thestudy of LANDSAT imagery. The volume of'the reports demonstrated thatan in-depth study of the utility of this imagery was made. It was notedthat the large, amount of information produced by the three PrincipalInvestigators was screened and combined into a single report. This isan excellent mechanism for cataloging the relevant results of the INPEprogram.' It effectively demonstrated the outstanding contribution ofthe Brazilian investigators to the LANDSAT program.

In his review of the Volume I, Chapter II, SEA RESOURCES (CF EmmanuelGama de Almeida) investigation, the Scientific Monitor stated that itwas a complete, well prepared and well organized report with an appro-priate format that adequately addressed each of the objectives outlinedin the original proposal. The report gives a good introduction to theinvestigation and touches lightly on the interaction of LANDSAT, Nimbus,Apollo-8, and Skylab video data products. Elaborate background informa-tion is supplied for each study area. Excellent use is made of trans-parent overlays on the LANDSAT images. The major detriment in thischapter is the lack of an introductory map of Brazil showing the locationsof :the five test sites.

Unfortunately, there are a number of content deficiencies that should becorrected before the Final Report is submitted. These are:

%

c. Page 27 to 30: Reference is made several- times to MSS channel4, 5, and 6-imagery but only a channel 5 image is included in the text.In fact, channel 5 imagery is duplicated on pages 24 and 28. Is this inerror?

* Page 45, first paragraph: Last sentence, ".1500 "m" should read"15.00 m".

Broderick to Dr. de Mendonca

2 ' • .

9 Page 45, third paragraph: Author refers to oceanic tides ofthe "daytime kind". Author probably means "diurnal - a tide having onlyone high water and one lov.- water each tidal day (every 24.84 hours)".-

e Pages 56, 67, and 59: Reference to F.ig. 11.20 and 11.22 is madeby stating the acquisition date of ths image.26 June 1973. Reference tothis illustration should bs by Fig. number. Near bottom of page 56,reference to this Fig. is 26 July 1973. One of these must be incorrect.

e Page 62, Paragraph II.3.1: The first sentence in this paragraphneeds clarification.

e Page 76, Paragraph II.3.5.3: Reference is made to a blowup ofGuanabara Bay showing river discharges. No LANDSAT image is included toshow this important result.

o Page 81,. third paragraph, line 3: "the isoline of 35% surfacesalinity" should be "isohaline of 35 0/00 "since salinity is measured inparts per thousand (0/CO).

o Page'88, third paragraph: Reference is made to items B, C, andD in Fig. 11.32 but not to item'A.

. •

In his review of the Volume I,' Chapter III, MINERAL RESOURCES (Dr. GilbertoAmaral) investigation, the Scientific Monitor stated that it representeda detailed and well illustrated report with an appropriate format thataddressed-most of the objectives outlined in the original proposal. Anobjective not addressed was the proposed study of the "main geotectonicunits of central and southern Brazil and their relation to known mineraldeposits". No discussion of the application of the structural data tothese mineral deposits was made. In addition, no information regardingthe "establishment of the best parameters for aircraft mission planningfor specific problems" was presented.

In addition, there are some content deficiencies that should be correctedbefore the Final Report is submitted. These are:

• Page 114, line 12: The sentence "The main differences..." is nota complete sentence.

«

• Page 114, line 22: "geologists who have"

• Page 115, line 8: -"implementation of software..."

t Page 117, line 19: "... has surveyed..."

Broderick to Dr. de Mendonca

3

9 Page 117, line 20: "... were spent for the .acquisition, of theimagery alone."

t

• Page 122, line 7: "... diagnosis..."

. 6 Page 123, line 9: "...maintenance..."

9 Page 123, line 11, 12: "costs about US $0.15 while the... havecost about..."

•e Page 123, line 16: "cost/effectiveness ratio... when comparedwith the SLAR ratio.

o Page 123, line 21: "Cambrian'age"

• Page 125, line 3: "...the previously available map."

e Page 126, line 11: "...geological community."

t Page 126, line 14: "...cost/effectiveness..."

o Page 125, line 18: Should read: "Only conventional photo-inter pretati OB techniques were used."

APPENDIX III

•t Page 13, line 15: "different"

0 Page 25, line 10: A period should follow "differentiation." Thenext word, "they" should be the first word of the following sentence.

• Page 25, line 19: "fine grained", "gray_waokes"

• . Page 26, line 2: "are mainly dates"

• Page 26, line 11: "argillaceous", "siltstone"

• Page 27, line 1: "Ladelra et al", "designated"%

• Page 28, line 7: "gray_wacke and arkosjc sandstone."

• Page 31, line 3: "...separated mainly by tonality;..."

. • Page 31, line 12: "...are jn^ unconformable contact with..."

O

Broderick to Dr. de Mendonca•

4

e Page 32, line 21: "It is unconformably overlying the unitsmentioned above." .

§ Page 33, line 1: "The area north of the Furnbs Dam_belongs to..."(no comma needed). .

• a Page 34, line 16: "thoroughly"

No reference is made to Plate 2 in the text. Also, the plates are referredto in a random order (Plate 3, Palte 1, Plate 1, Plate 4, etc.).Additionally, the plates should follow immediately after the first refer-ence to them, rather than being grouped together at the end.

Map 1 should follow after the first reference to it on page 20. Likewise,Map 2 should foilow- directly after on page 31.

In his review of Volume II, Chapter IV, SOIL RESOURCES (Dr. Fernandode Mendonca) the Scientific Monitor stated that this report completed theanalyses relevant to the stated objectives of the investigation. Theinvestigation went a long way in identifying promising and economicalapplications of remotely sensed earth resources data in the field ofagronomy. Technical details in the form of LANDSAT imagingjllustrations,overlays, and maps v/ere amply provided. In addition, the usefulness ofLANDSAT imagery in other discipline areas of land use and demographicstudies was explored. Unfortunately, the study is labeled as preliminarybecause of the absence of ground truth.data.

The content deficiencies that should be corrected before the Final Report issubmitted are: .

• .

"t Page 141, Paragraph IV.2.4.3.6: This'statement needs clarification.

• Page 145, Paragraph IV.2.5.3.4: This statement needs clarification.

• Page 149,"Fig. IV.7: Only 4 area types are described in theLegend although 6 area types are discussed in the accompanying text.

• Page 150, Fig. IV.8: No referjnffce is made to this Fig. in.the text. • '

»

t Page 151, Paragraph IV.2.6.3.2: This statement needs clarification.

• Page 155, Paragraph IV.2.7.2.3: There is a discussion of theParana River but it is not designated in Fig. IV.9.

• Page 161, Paragraph IV.2.8.2.5: This statement needs clarification.

&' O I

Broderick to Dr. de Hendonca

0 Page 177, line 16: Refernece to Fig. IV.20 should beFig. IV.21.

o Page 179, Section IV.3.2.2.1 Legend: It is unfortunate thatthe terms "cerrado", "cerradab", etc. were not defined earlier in thetext. .

Will you kindly consider incorporating these comments into your draftbefore submitting eight (8) .copies of the Final Report to:

Goddard Space Flight CenterGreenbelt, Maryland 20771Attn: Scientific Investigation Support

Code 902.6

Sincerely, .

M Ctames C. Broderick

\NDSAT Technical Monitor

Documents

E7.6- 10.0,7 a - NASA · RECEIVED NASA STI ACQ. DEC041975 E7.6- 10.0,7 a SEPLAN-CNPq Institute de Pesquisas Espaciais - INPE Sao Jose dos Campos - S.Paulo - Brasil (E76-1CC73) CCILEC-IICN